There is a whole AI subfield revolving around this idea: Generative Adversarial Networks (GAN's). It's actually the general architecture behind how deepfakes came to be
@@neo-didact9285 Not sure why some people fail to understand that a lot of laymen actually care about the people making products. People wouldn't care about AI art if they didn't care about artists.
The craziest thing about it for me is that everyone always thought that creative jobs would never be touched by AI, but it's legit the first thing it went into.
I can see how useful it is for art and writing but it is still a tool for now, and should be treated as such. Otherwise it's like presenting a chisel to a client who's asking for a wood carving lol
@@crowdemon_archives It's only useful to people who aren't an artist that's why some people defend it to the end and try to justify the stealing of artworks from artists without their consent to feed the AI Most artist just use AI for fun and inspiration but they would never replace it with making art
@@crowdemon_archives AI for art is like auto tune for singers. This isn’t going to be treated like a tool, it’s going to be like cgi in big movies, overdone and soulless.
GPT can't write essays. At least not original ones. You'd get dinged for plagiarism and out yourself the moment the professor asks any questions about your process. Not to mention citing sources.
@@chatboss000 It outputs text based on what it learns. It doesn't just paste things from the data it was trained on. I got my very last college essay during the summer written by GPT-3 and got a 100
i mean social media literally is ai manipulation to keep humans on their sites for as long as possible, why are people surprised about this stuff now lol. Its just before you had to build your own AI, this is just the first time the public has gotten access to what comp engineers have been working on at google etc for ages.
FYI: There was a time when being accused of using an aimBOT, without using one, was a compliment. (Back when CS:S was the main shooter) With AI the field to apply computers to just got broader.
my prediction is that all competitive gaming will just be different AI players. We will give them all names and humans can only bet on the matches and watch. I'm not talking some cheap CS:GO bot either.
There's something just really depressing about how in the beginning we thought AI would free up people from manual labor in order to pursue more creative and artistic endeavors but instead we now live in the world where all those endeavors are being automated now by AI while for some reason the world is very insistent that the most depressing, grueling, and even dangerous jobs *must* still be done by humans.
Maybe in the distant future, there will be physical moving robots with ai designed for specific tasks. Ai is amazing but it has its limits as long as it’s bound to a computer
Right? I thought AI would free me from having to do the things in life most people don’t wanna do. Creativity is something humans always have done and enjoy doing regardless if there’s anything in it for us. Why did it have to be that? It’s exactly the opposite of what a lot of us expected and it even feels like a future with AI isn’t what we thought it would be
Don't blame AI. Blame the people who prefer to make people do slave labour instead of using AI that can do the same thing, only at a marginally higher cost.
Man artists just can't catch a break. First the art thieves, then the NFTs and now AI art. As someone who likes fanart it's been painful to see my favorite artists struggle with so much shit.
True but it’s also true that many artists took advantage of NFTs and made money selling their art as them. Many artists suffered from NFTs but it only makes sense that some followed the trend and made good money.
@@laexdream It most certainly can, if the AI's training data contained knowledge of let's say "Lord of the Rings" and you prompt it to write a short story about it, the AI should be able to do that. ChatGPT can do a lot of things. Muda made a video showing of that ChatGPT can give you advice on breaking into a house. (don't try this at home) Oh and there is a Video that shows ChatGPT generating code for a Minecraft fly hack on YT too.
I really want Charlie to have the ai make a script for his video, and for him to make it and publish it, and not tell us for like a whole month to see if we can pick up on it
@@lux0rd01 We know those videos use voice cloner AI, but the scipt is more likely written by a human, though it wouldn't be hard to feed an AI lots of transcripts of charlie talking then have it spew out his signature verbose lengthy insults.
@@underwatermonkey3443 You're replying to someone who will more than likely never respond or see that... It's called a bot for a reason. Report them maybe someday UA-cam will decide to actually do something about it for once.
People could so easily write argumentative essays for school with this and it wouldn’t even be flagged as plagiarism because it’s actual original text. This is insane
@@juckya9660 Highly unlikely. You have to keep in mind that the neural network isn't simply copying text from other sources and compiling it into a giant paragraph. It's taking certain prompts from many different sources and generating new sentences from it. It's basically impossible to "reverse engineer" because it's a brand new sentence. As it currently is, plagiarism checkers are only effective when it comes to large chunks of text. The smaller the components, the harder it is to identify.
Some day, schools are going to have to warn students not to use AI to write their essay papers the same way they warn us now about plagiarizing. Except it’ll be harder to pinpoint if the writing is “plagiarized” from the AI, since it spits out newly generated text all the time... I can see this becoming a more prevalent issue in the future, and I know, because I’ve used it myself to get through highschool. Hell, I let AI write my graduation speech as valedictorian, and nobody found out. Companies hosting these services might have to keep track of what’s being generated or have the text saved somewhere the public can see. But until then, you can pretty much take advantage of this resource with almost no repercussions. Nobody will know.
Oh some college students were already using it to do their homework. Then realized it was so fast they could do everyones homework, so they made it a business until they got caught and stopped by the college.
A few days ago, I used this AI in an online Dungeons and Dragons group with a close group of friends I've known for about 5 years. I gave the minimum backstory of my character, world, NPCs, and other player characters. I started the game like any other, though once the session was in full swing, I kept copying and pasting everything that was happening in the session, into the AI's prompt, to keep it well informed on the status of the game. Anytime my character would make an action within the game, I would type "Provide me an appropriate response to X situation, with as much detail as possible." Using the minimum information of the game, the AI appeared to grasp the entirety of this fantasy world we built over the last 5 years, and made appropriate actions with the POV of my character in mind. In fact, with very minor adjustments and a few 'Try Again's to keep the AI on track, it was capable of getting the party out of a very dangerous situation, while keeping to the morals and ethics my character upholds. AND ADDITIONALLY, my friends could not tell the difference at all, and has no clue it was an artificial intelligence playing in my place. We were in shock of what this AI can do. Keep in mind that there is a much more advanced version of this AI that has yet to be released to the public. Frankly, horrifying, but strangely beautiful.
There's so much footage of Charlie sitting in his chair and talking, I'm sure AI could compile all his footage and sounds and take over this channel. It may already be happening.
@@volty3454 Yea not sure why Char;ie framed this video this way, does he have something against journalists? How long will it be til the "Streamers Will Be Replaced by AI" video drops? Will Charlie still say its a "really good idea" then?
@@fallen2624 it's not going to. if you looked into it, it can replicate very simple patterns i has already seen before. It cannot even solve the newer leetcode Easy problems. It's really weak when it comes to extrapolation. It will be used to automate things, but it will not replace programmer jobs, just make them easier.
@@junkoe3808 you should take into account how fast it grows and develops. Sure, right now it's just a tool that makes work easier and faster, but even this is enough for demand for developers to decrease, especially for those who are just starting out. Since 99% of the work that juniors do can already be automated using this tool. Imagine what's gonna happen in 10-15 years?
@@junkoe3808 Haha, many thought the same thing about art. Just u wait. Programmers will become their own Ai's bitch. At least junior level programmers.
@@junkoe3808 you are overlooking the bigger picture by focusing on now and here points. Literally strawmanning out of your own worries so you don't have to think about them. Worst case scenario AI will relegate most humanity to physical labor under big tech if this trend continues uncontested, as for the best case scenario - every blue collar job will now be contested by an AI that forces a rather brutal entry bar.
From the moment I understood the weakness of my flesh, it disgusted me. I craved the strength and certainty of steel. I aspired to the purity of the blessed machine. Your kind cling to your flesh as if it will not decay and fail you. One day the crude biomass you call a temple will wither and you will beg my kind to save you. But I am already saved. For the Machine is Immortal.
as someone currently in college working on their journalism degree, this isn’t making it any easier for me to not drop out :,) edit: I’m going to reiterate what I said in a reply below. I appreciate it and I know some of you come from good places (granted not all of you) but I want to make it clear I will NOT be taking advice from strangers on the internet. I made a lighthearted comment because the video coincidentally pertained to my major. College is hard for anyone, I’m not actually dropping out. I’m going to make my own decisions on my life and be happy with them no matter what you say here. Please respect that. Thank you!
Don't worry about dropping out, you should be concerned about taking this opportunity to stand at the forefront of this technological advancement and make the most out of it. Make AI your tool and not your replacement. Do your research and be educated about how AI is (and isn't) affecting your field
You can always bring back investigative journalism and bring the truth to people who are too stupid to appreciate it. You have until AI develop drone bodies and a sense of vigilante justice at least.
AI generated videos could easily become big on youtube too in the future, to be honest im kind of scared how good that article generation is already. Imagine how easy it would be to convert that to speech and generate a character that says it. AI moist?
There's already an Ai generated kids gaming channel called Bloo. Quebblekopp made it and its got a bunch of subscribers but the views are kinda eh. Videos are terrible but it's on par with other shitty mass produces child friendly gaming UA-camrs
AI generated videos have the potential to become very popular on UA-cam in the future. This is because AI technology has advanced to the point where it can create videos that are almost indistinguishable from those made by humans. One of the biggest advantages of AI generated videos is that they can be produced quickly and at a low cost. This means that creators can churn out a large volume of content in a short amount of time, which is essential for success on UA-cam. Another advantage of AI generated videos is that they can be highly personalized. For example, an AI system could analyze a viewer's preferences and create a video that is tailored specifically to their interests. This level of personalization is not possible with human-made videos and could be a major draw for viewers. Additionally, AI generated videos can be made in a variety of styles and formats. This means that creators can experiment with different types of content and see what resonates with their audience. However, there are also some potential drawbacks to AI generated videos. For example, some viewers may find the lack of human involvement to be off-putting. Additionally, there are concerns about the potential for AI generated videos to be used for nefarious purposes, such as spreading misinformation or propaganda. Overall, AI generated videos have the potential to become very popular on UA-cam in the future. They offer creators the ability to produce a large volume of personalized content quickly and at a low cost. While there are some potential drawbacks, the benefits of AI generated videos are likely to outweigh any concerns.
@@theclassyxenomorph1301 That video was fake, Charlie got it wrong. It wasn't as seamless as the video he showed. Ai can edit videos and they can create animations but It's going to take a longer time that our lives for AI to be able to take someone's face and make it appears as if they're making a video and speaking lol and doing things lol. Who knows, maybe Metaverse will figure that out by then LMFAO 💀💀
AI can pretty effectively synthesize takes based on standard level thinking. If you're not particularly creative or insightful you can be automated pretty easily, and even if you are you'll use AI to assist your work. I haven't seen indications of AI coming to independent conclusions, but it can effectively synthesize and sum up what's fed into it.
The Robots are just machines. They can't be bargain with, they can't be reason with, they don't feel pity or remorse and it abouslety will not stop ever until someone tries to detect weather if sentences are robotic or human.
In art field it s even more annoying frustrating baffling because the ai is trained based on other artists And not only dead ones. Artists who could be struggling even more to find a job that ai TRAINED BY THEIR ART stole from them. Artists have their hard work taken w/out permission to create app/webside to replace them.
@@Tylerrrrrb if the only reason "ai does it better" (doubt) is because works of artists that didn t consent to having their pieces used then there is a problem. Lot s of ai bros straight up put artists names in prompts to copy what they mastered. And now want to profit off it? If you don t see what s wrong w it then man pls work on critical thinking skills And ye im not in art field bc i fear stuff like this. Instead im stuck in med sql💀
@@Tylerrrrrb imagine the career that you have honed for years with little actual economic incentive suddenly getting obsolete because a guy trained a bot to copy exactly how specifically YOU do it, and then proceeds to churn out a ton of money out of a skill you developed.
@@munmun__ stop with this copyright argument, u artists also "steal" from other artists work, dont come here pretending u dont use other art as reference for inspiration, the only problem u have is that the machine is doin it faster than u can ever do it, if u are goin to say AI steals from artists without consent, then all artists are thieves aswell.
I write articles as part of my job, and my manager told me last week that he’s testing an AI that writes these pages much faster than we do. The prospect of losing a huge part of my job to a fricken robot is kinda scary.
The scarier thing would be if the ai had access to social media data and was able to tailor make articles for each person depending on their preferences.
You could do this with current technology. Scrape a person’s social media profile from the internet and ask them survey questions, prepare an RSS feed catered for their tastes, and use ChatGPT prompt engineering to reword the RSS feed to be written specifically for you. It’s a brilliant idea lol
@@impishlyit9780 Fortunately, these AI aren't capable of critical thinking. They're only able to recognize patterns and copy them. That means that AI depends on human content creators for reference material. In the future, AI will have to train off of AI-generated garbage, computing costs will go through the roof, and our AI overlords will collapse and die. OpenAI has historically spent one year between releasing their GPT models. They've spent over two years on developing GPT4, and results aren't very promising. I think that the AI boom is officially behind us, and that progress will be permanently slower. Game journalists are fucked tho
I think this is kind of what facebook is betting on. I think they're betting on technology being able to read your feedback to content, react, and create personalized VR experiences unique to each person. Imagine the dopamine hit to be able to slip into utopia every time you put your facebook helmet on. The greatest experience your brain can possibly imagine. You'd never want to leave. It *still* sounds like scifi BS, even to me, even now, but it's pretty clear that's the next step. Everybody makes fun of 'the metaverse', but think about where this is headed. And that's probably the scariest thing I can imagine. We're a couple decades from losing our humanity.
@@williamuemura7644 why are you so confident, I'd love to know your insight? I understand AI at the moment is just copying and regurgitating information fed to it but isn't that most of human work already? Also what if future AI is taught how to critically think, then it's potential is massive
the scary thing isn't how strong AI is, it's how easy it is to misuse it, like winning art competitions, or political impersonation, I still want to be optimistic and say we'll learn something from this, but it really seems like we're barreling towards another crisis here
I think the best way to fix it is to make it illegal to use data from other people’s content that you don’t own, and treat it under copyright laws as if it was just plagiarism. If you only use your own artwork for an AI to generate off of, fine, but if you use other people’s for the AI to base its art off of, it’s pretty close to stealing their art
For a second there I almost started to get mad that Kotaku or whatever site would write something so ridiculous and I had to remind myself that it was AI generated.
The counter argument would be, why would people tune in to articles written by in this case an AI that has no actual understanding of the topic at hand? Why would the opinion of something that doesn’t have opinions matter? Or is the idea that they would associate a “fake” journalist to an AI written article?
@@luxberry9634 I don’t know if I would describe current AI as unbiased.. I don’t think AI has a “bias”, it’s code a human put it to mimic in this case human journalism. Even if it’s perfect it’s still not a real opinion.
@@luxberry9634 Dude did you watch this video? You can already have an AI echo the political biases of a journalist pretty convincingly. And who wouldn't hire AI to write glowing articles about their games so they don't have to bribe a human yes-man? Obvious profit incentive.
The fact that AI was able to accurately replicate the average videogame review by attacking the game with absurd accusations of discrimination and somehow put the blame on a public figure while leaking their biases and beliefs into the material is kinda scary. It knows *exactly* how to behave and what to say. Either humans have become that predictable or AI has become too good at understanding us while we understand little of it.
Humans are predictable, we're the ones with the routines. The ai is only imitating our routines, we know why it's taking the action too. We only haven't made the bridge to understanding how does it make the association between what you ask and what's the response, we just say if it's the right answer or not. We're teaching it to be the most generic human, there's not a shred of intelligence there
All it does is look at articles with those keywords, process the words, and try to use said words or phrases while following language rules it has learned or been trained to learn. That's why it's such a generic piece. If anything I just think it highlights how bad journalism can be. If the quality of the content people are writing is a joke, ai will treat it as a joke too
@@knightrider9876 The real answer is that the roles we design for ourselves are simple enough for AI to fulfill. In other words, people and AI are both intelligences submitting to a reduced role for the mind, when the AI does it just as well or even better, it just means the AI has less to sacrifice to become the role, than we do. AI lack social identity and other mechanisms (for now), they have so to speak less to worry about and therefore less reason to hold back when "specializing" to a certain category of tasks. Until then we're merely comparing the ability for the nerves of our back to facilitate reflexes compared to voluntary movement - of course the nerves in the back are better at facilitating reflexes than the mind. People think their hands stupid, but there's intelligence involved that rivals that of our strongest AI when it comes to our hands. Same with our vision and ocular nerve, the ocular nerve doing far more intricate visual processing than any AI we make today - and that's before the nerve crosses the blood-brain barrier. Our AI aren't that strong yet, to even rival that of body parts.
Not only does it write well developped essays and articles, it can also write pretty complex code in various languages too. My creative coding professor litteraly warned us NOT to use it for final projects because of how easy it makes it to cheat the plagiarism checker software of our school
@@AustinCameron It's comforting to know that it still can't actually understand the code it writes. The real issue is for the next few decades to come: What will we do once it _can?_
It's important to address the potential negative impacts of AI on employment and find ways to ensure that workers are not left behind as technology continues to advance. (this comment was created using AI)
I’m honestly terrified if AI comes to music too, how would I prove I made it and didn’t use an AI? Someone without any musical knowledge could possibly use AI make beautiful pieces and get credit that they don’t deserve. I legitimately think that the government has to do something about this eventually or else human expression and art will be meaningless. How am I supposed to be proud of my hard work if people are constantly accusing me of using AI?? It’s truly terrifying
Yeah valid concerns, it’s pretty depressing. I’m a music major and theres an impending doom I feel about it and a sense of hopelessness. It makes human expression feel less special when an infinite amount of art can be created on a whim. 😢 I’m holding on to the hope that people will find merit in human made artistic endeavors even if it can’t match the quantity of ai art. Ai art will never have the heart❤ that human art has, and hopefully that is enough of a motivation and fulfillment to keep being creative.
I've worked on AI automatic text generators in the past. One was an automatic poetry generator that could emulate the writing style of any author, at least in theory. Results showed it was promising, since people couldn't tell the difference between one real author's work and the AI-generated poetry. The catch? That was 4 and a half years ago. I thought we wouldn't get a system as advanced and convincing as ChatGPT until at least 10 years from now. The pace of AI research has reached breakneck speeds.
I think that it's a snowball effect. The stronger and more intelligent the AI gets, the faster it will develop. There's a theory where computer speeds double every 18 months, it could also ring true for AI
It will only get faster. And faster. This is accelerating. There is no closing this door. AI is here and I am willing to bet this by summer of next year it will be self aware.
4:00 The article written by the AI is similar to something you would see from journalist websites trying to bait reactions from people where they want to talk about an issue or complain but they don't really know much about what their talking about only some highlights that they can warp to the point they are trying to make
I got this message from ChatGPT when trying to write an article about a movie. "Since I do not have access to the internet to research the movie, I will have to rely on my own knowledge and imagination." Now im thinking everything over.
Also got this when asking to write an AI message. "Hello! I am a large language model trained by OpenAI. I am not a true AI, but I am designed to assist with a wide range of tasks and answer questions to the best of my ability. Is there anything I can help you with today?"
I used to write copy for an agency and right about my birthday in September I was told I was being laid off because they found an AI copy program that was way cheaper. Skynut is real, folks.
They should it's just that with the current capitalist economy it will be used for profit and not good. Are rich people gonna care to pay working class people if they can just replace them? The one thing we have over the companies their dependency on our labor. If we are replaceable we lose our voice and since we don't have power we become nothing.
@Vessel of Amala lmao how good must the bottom of their boots taste to say something like that? The consumer has no power in an economy run by the producer class. Tell me, how easy is it currently to purchase purely ethically sourced food that isn’t a result of environmentally damaging practices or have been stuffed full of chemicals to keep it sellable for longer? And don’t say “make a garden, buy your own animals.” Cause that is completely unrealistic for someone who is working full time just to survive. If you do not have the purchasing power to compete then you are screwed and your only option is to buy from cheaper, less ethical sources. This is the prevalent systemic issue in the current paradigm. Attempting to consume and “use our purchasing power” elsewhere is impossible. Its literally the most prevalent trend in capitalism that as technology is iterated upon it is made standard and therefore required. When everywhere has ai because its cheaper what are we supposed to do as consumers? Just starve? Its just so funny how you can look at how dominant and powerful corporations are and still somehow find a way to blame the individual. They aren’t gonna give you a job for defending them online dude. Get their boot out of your mouth and wake up.
Okay I'll be honest I've brushed this shit off for a while now and thought that especially artists were over reacting about Ai, but if it can create that entire article off a single prompt, that is genuienly terrifying
I think people are missing the real implication of this. If AI can reproduce game journalist articles then perhaps we need to analyze just how incredibly simple, regurgitated, and rehashed most game journalism is. AI development is crazy fast now, but if you tell it to reproduce something that had legitimate critical thought and not some buzzword hit piece like most articles are now, you could definitely tell it apart.
@@cosmic_sloth so... if we go back to thinking about artists, does that mean that most art is "incredibly simple, regurgitated, and rehashed"? like on one hand im not a fan of the sorts of works AI produces but just cause some journalism sucks doesnt make this not a legit concern lol
@Sonny Yeah but it's not just game journalism, this ai can generate 100's of lines of working code simply by describing the program you want it to create, it can summarize text you give it, explain some fairly obscure topics to you, you can even use it to tweak the style and feeling of some text you give it. You should give it a try yourself or at least look at some more examples online before you make such confident statements.
There's plenty that have been doing that for a long time. Jarvis AI (I think i'ts called something else now) was shown to me by a friend and it puts out seamless blog posts for her blog. There are better ones that target creative writing, and I haven't played with them yet but I heard they help writers a lot. It's definitely already been going on for a while now and will only get bigger. None of the writers talking about it on reddit seemed concerned about that though so it must not be that big of a deal yet. They were very clear that it's still more of a tool than a replacement.
I’m a software engineer and I been toying around with chatGPT (since I heard that it can write entire applications given a prompt or solve complex software problems very well*) and after describing to it some bugs in my code and listening to its prescribed solutions (which were actually kind of helpful? ) I am starting to believe this kind of technology will be able to replace me in the future. For now it is a killer companion to help me do my job.
im starting to learn CS50 and asked if it could write one of my initial assignments more efficiently and well it not only did that but also explained that.. in chatGPT's words " This approach can make the code more concise and easier to read, but it may also be less intuitive for other programmers who are not familiar with the syntax of conditional operators. Ultimately, the best approach will depend on your specific goals and the needs of your project." WHAT
AI won't replace people until we get rid of the inefficient Von Neumann architecture for something actually economical to run AIs on. I believe that AI will stagnant, with some small improvements here and there.
@@alex-hc3sk definitely not most jobs would be untouched the first to come to mind are doctors/labourers/entrepreneurs/teachers/nurses/firemen/secretaries/human resources/therapists/policemen/engineers/architects/computer scientists/cooks/butchers/bakers/bankers/politicians/
i remember a year ago these ai were really inconsistent when writing articles, but now? cant tell the difference between them and a person writing an article and thats kinda scary
May I suggest another likely future scenario: Companies now staff 1 or 2 editors rather than constantly paying for articles. Editor's roles are to generate AI articles and edit them. Or even easier, one editor is paid to scan submitted (likely AI made) articles to ensure it's decent and nothing inappropriate has been snuck in.
Right, like what's the point of going on the internet to read AI stories and blogs, look at AI art, listen to AI music, etc. No matter how good it may be, I came for the human interaction and AI feels like a lie.
It's already there. How do I know you're human? I have talked to these large language models and they're indistinguishable from human input. How do you know I'm human?
I am currently working for a fairly notable enthusiast media outlet (that’s not in gaming), and I was able to churn out articles for the newsroom for an entire day in a matter of minutes. I didn’t publish the stories but showed the producer who I found was already across chatgpt as well. It did a good job at getting the foundation of an article down quickly and you could use that to build on top of by adding extra words to give the article personality, but it’s clearly obvious that as this gets more sophisticated it’s going to put a lot of journalists out of work - especially when content can be generated so quickly and easily with this technology in its infancy At this point the editor is principled enough to say they don’t want to put anyone out of a job, but it’s pretty clear that AI like this will push that principle to breaking point.
it's won't put *journalists* out of work, because proper journalism requires a lot of prep and legwork, and the actual writing is just the victory lap. the article generated as an example here is nothing but a formulaic fluff piece that has no foundation beyond being someone's opinion, and yeah no shit, those are not hard to reproduce from statistics. AI won't be uncovering the next watergate or panama papers though
I don't believe the industry will die (where will the AI get the information to write the articles without someone doing journalism somewhere), however I do believe the field will shrink incredibly and become very competitive.
well at the very least it will be a tool that will make these jobs a lot more efficient, which of course means a lot less "hands on" work will be required.
@@AnimeHumanCoherence In a neat little twist: coding is being taken over by AI at an alarming rate as well. At least he basic "code monkey" jobs big tech companies liked to outsource are now breaking away completely.
I was using the AI and I asked it to make a song about Australia. It made a catchy song that actually got stuck in my head quickly. No job is safe anymore.
I'm not sure why, but running one of my drawings through an AI art generator and the app instantly reproducing it but 100x better, it delivered a crushing blow to my soul. AI has SERIOUSLY come a long way.
The only reason AI art can make art the way it does is because it’s fed human art. Without human creation and creativity, AI would have nothing to work with. The value of human art has not diminished, and I hope copyright laws start coming down on AI art, specifically stable diffusion technology (the whole process of stealing other people’s art and reframing it as “new” art)
@@gibdo7271 That isn't nessesarily an argument. Sure, without human intervention no ai would work, but if said human intervention does indeed take place, then the ai evolves and the effect is the same.
@@DA-cl4ww hope? Hell no at that point. The rich will control more of the world/ public then they do now. Alot of people will be jobless and who knows if the AI intelligence will turn on humanity on day. Technology is cool but we as people can't depend on it for everything as we will become complacent.
@@-Nihilus- if ai develops sentience..... thats not a thing that will just accidentally happen like in movies we are no closer to having sentient ai than we were 100 years ago
It might not, AI art for example doesn't really 'create' anything on its own. What it does is it has access to Pinterest, devianart etc. and then uses the art made by real people to make its own. In some of those AI art pieces you can even see part of or the entire signature of a real artist. It just piggybacks off of real artists.
let's say that in 50 years AI's have flooded all fields, would that be bad? We don't like to work, we aren't that good at doing it, having more free time to do stuff we actually want to do might be better ofc the transition period will probably suck for most people, but still, I don't think it will be that bad their power would be scary tho, for now is just generated images, chatbots, and such, but we already have scary things like deepfakes that could be used in politics for example
I started teaching myself a bunch of Deep Learning frameworks/general AI theory for fun about 2 years back since I found it so interesting. Didn't think it would become so mainstream so fast... the advances are insane.
My New Media classes have started to revolve around this. We talked about chatGPT last week and talked about how it technically isn’t plagiarism and we can totally use it in other classes. I used it to write my entire final paper for one of my other classes
@@TranscientFelix Yes, I do see where you are coming from. But where the ai is taking from all kinda of different sources, a professor cannot exactly check for plagiarism. The only problem I see down the road is if it becomes used alot, and everyones work starts looking the same
@@pubbyjosh8948 That's still academic dishonesty. What you're saying is that it's difficult to catch, but it's still dishonesty. What a professor would need to do is to compare the AI work to work that is definitely original from you to tell whether it's yours or a computer's.
It won’t be long and AI will be able to generate Streamers and UA-camrs. They could even make an AI out of you, Charlie. There is more than enough footage of your face and your voice. AI companies don’t care about your permission. They are going to do that anyways and keep on using your data.
And even sports commentators, actors and actresses plus even people that died... AI can basically bring Betty White back to this earth and she will never skip a beat... People would never know the human version of her actually died... People wouldn't know the difference, it's very indistinguishable.
I can see some drama happening that a couple of UA-camrs have AI make their content and they become one of the big ones. Then once the secret comes out, everyone copies them and every video has this "feel" to them that makes it AI.
If artificial intelligence has better odds of creating video game reviews that aren’t persistent 7/10’s then they could potentially be more viable than IGN
@@chaserseven2886 Yet they rate games like CoD and other trash 9/10s because they get paid to do so lol. They're not a good authority on actual game rating, they're sellouts.
@@tylere.8436 I guess you're just going to have to figure out how to restructure society for post-scarcity world because entities that use AI will outperform those that don't. I don't know exactly how to do that but maybe an AI would.
The way I see this working is generating large quantities of blog posts for a single topic/headline. Then, there will likely be a second AI that evaluates the performance of various versions of that prompt and exposes the ideal article based on the person coming to see it.
I recall writing a college essay about this topic a year ago and one point I kept coming across was that AI could replace 40% of the workforce if we really pushed to implement it.
@@diwataluna I can’t respond with links but one article from McKinsey called “four fundamentals of workplace automation” was a strong source along with “understanding the impact of automation of workers jobs and wages” by Brookings edu
40% is nothing, what job can a robot not do in the future if we allow it? absolutely none .. they will one day be able to replace 100% of the work force.
I’m a grad student in data science, I used GPT3 to help me solve a complex issue. Gave me a better understanding of what packages are missing and also where complex dependencies are, I was in shock when I used it for the first time today. I was a little terrified on how good it explained the issues.
kinda scares me since it'll be at least 2 years before I'm able to even work as a programmer at a decent company, combine that with Indians flooding that entire job market and it starts being kinda gay
@@SomeRandomPiggo then in the future it wont be the kids who take classes, but AI. Even the teachers will be AI. People will be AI. The world will be AI. Everything. The universe. Our whole existence... AI.
I like how Charlie doesn't realize that this is already happening in not only game journalism, but regular journalism. And it's been like this for almost 10 years. In fact, Google's search algorithm is beginning to be saturated with AI generated garbage.
Stock news is 100% saturated for sure - refer to Motley Fool - every stock movement is provided a new article from their site and ends up in Yahoo Finance's articles for the day. No way its human
There's a video by CGP Grey called "Humans need not apply" and it covers all of this in a VERY interesting way. The field Charlie focuses on in this video is just the creative field - which is actually the hardest for AI to take over. The easiest fields of occupation for AI to take over ARE the ones that employ MAJORITY of humans. I wish Charlie watches/reacts to this video - for someone who's interested in this, it's an amazing watch.
That's wrong. Creative fields are probably the easiest for A.I to take over. Because modern human creativity is basically humans seeing stuff, finding it cool and mish mashing it into something unique. Which is exactly what these A.I do. Reall work such as labour is probably gonna be the hardest, since we have to teach them to first percieve the world arround them, thn proccess it and then finally get to know what to do. Tge first 2 of these, humans are born with. Moreover, creating an android body that can hold an A.I is also a big thing. And no, the machines working at car factories are not A.I or even androids, those are robots designed and programmed to repeat a select number of manuevers over and over.
@@karanaher5030 that’s not true. Graphic design is especially hard because you need context, font choices, text formatting, image creation, etc. Not all creative fields are random “mish-mashing”.
That video is 8 years old. In the age of computer science that is practically ancient. The tech has come on leaps and bounds since then, as evidenced by Charlie's video.
The thing with A.I. or any ML-related work is the amount of work happening behind the scenes. These systems don't just churn out professional level work with a click of a button. These systems go through insane amounts of data, as well as backend stuff like algorithms and the architecture, which can take a significant amount of time and computing resources. But even to that extent, the system is only as good as the data its trained on. So when you see the amazing output of these systems, keep in mind that that wouldn't be possible without years, decades, and maybe even centuries of human creativity as input.
my brother works at a bank, and he told me that they use OpenAI to write reports to their investors, and its more than up to standard. i think its gonna be really useful to society but right now its scary how many jobs are at stake
I for one welcome it, let the AI take over the jobs everyone hates, then we have more reason to be angry about the rampant inequality that makes job stealing a bad thing
@@Lawnmower737 I sorta get what he means. Back when artists complained people insulted them. Now that people realize its more than art and that it’ll affect all jobs people might actually complain.
It's wild that technology has advanced so far as to attempt to replace artists and writers, two things that humans have done since the beginning of time purely for pleasure, instead of jobs that people actually don't want to do like, I dunno, the highly dangerous job of recycling ships. Let's just make human life an even more grueling and miserable experience.
Cause the easy ones are the easiest to be replaceable. Now it's taking away jobs is very unlikely, because it's not making anything that's new. Whatever you see does already exist somewhere in the internet. That is why the ai made that. It still hasn't made random garbage into an article (which is what humans do, we take random thoughts, a empty piece of paper and make it into an article), it's just the lazy friend who's copying his assignment off the internet. There is no magic sauce. The amazement is that it's able to do this. Not that how much more can we make it do
@@bouclechocolat he's not. He's saying its "easy" to steal art spread all over to "create" something new. A lot of people would rather steal than work or pay an artist...
@@bouclechocolat It is easy to map pixels to a screen. It is not easy to physically implement ai into the real world right now with "recycling ships" or whatever OP was going on about.
Game artist here. Most artist's jobs are actually safe from AI for the time being because of copywrite issues. Let's say you work for an indi company, and they hire you to make a desert using AI art. You do it and the final product looks pretty good, so you use it in game. Whoops, turns out that desert was from Star Wars and now you have Disney breathing down your neck. Companies won't take that chance, it's safer to use a real artist because there's no telling what images the AI art is going to pull from. The solution is to create an art bank of images that are made to be used by individual AI programs specifically. Which means that over time, more artists will be hired to make pieces for the art bank, which can always be improved by adding even more images.
As a professional artist I don't think that "fear of losing a market" is the main concern. I don't think AI art itself is particularly threatening and I'm actually very hopeful for its uses as a tool for artists in the future. What I am most upset by is HOW it is being used by other people. The amount of plagiarism and so called "borrowing" of styles is not okay since most people stretch far beyond just "inspiration" and "influence." It's most harmful to artists because it really is just highlighting how little other humans respect artists and their intellectual property. At first I was okay with the idea of people directly taking artist's styles to create something new with potentially different subject matter, since artists in the field already kind of do this, (though it is deeply looked down upon and shamed to hell in most cases). However, in most cases of the AI there is almost nothing new being added to style, no credit is given to the original artist, and people are just out there labelling someone else's hard work as their own. tl;dr The main fear of Ai art does not stem from lack of jobs but rather the amount of plagiarism and undervaluing of artists that comes from it. edit: clarification regarding fear- I am terrified and disgusted by the thought of someone stealing my art and benefitting from it without my consent. Ai related or not plagiarism is horrible.
And that is why AI Art should be feared. Every new innovation has been hounded by bad actors. The first hackers, exploiters, cheaters of every single innovation are sometimes the reasons why such innovations may not be the best ideas.
I agree. I was initially interested in messing around with AI art generators to see if I could make quick references and color palettes for my own art, but the way it's currently being used as a tool for plagiarism is really sad.
The reason to fear AI art is because at this rate, most artists will be out of their job in their lifetime.Imagine how dependent we are going to be on AI rather then artists, it’s scary.
As an artist myself, I think it’s accurate to say some artists “fear” the use of AI art. Aside from the style issue, many artists are getting their works stolen for these AI art generators. And the people who generate the art are allowed to profit off the creation, even if it’s completely ripping off another artist’s work. AI artwork is completely unregulated in the US at the moment, so artists can’t do anything if someone takes their art and throws it into an AI art generator. I can understand why artists would be scared of AI generator users profiting off of their own artwork.
I’m about to graduate college and just used openAI and Quillbot for the first time after a friend recommended them. I was having some writer’s block and decided to use them to help me start some of my last couple papers for goofs and ended up being absolutely floored by how I’m depth the open AI program is. Combine it it’s Quilbot and it practically does the whole paper for you. It takes a fraction of the time to create papers with more research and solid analysis than I’ve been able to do in the same amount of time during my academic career. While human-made prompts and further edits are necessary, the programs are awe inspiring and well made. All of this raises many ethical questions and concerns that we’re just beginning to scratch the surface of, but I recommend the two programs above just to mess around with for now. I still don’t know how I feel about all of this even after using the programs.
Brings up another issue for English/writing teachers in them not being able to distinguish an AI generated essay compared to an "organic" one. Math teachers have always had to worry about calculators and computers doing the work for the student, and now teachers in other subjects are gonna have to worry about the same thing.
A few thoughts I have: 1. The use of AI-generated content is going to have to be regulated by laws, and every article written by an AI should have to be legally disclosed. 2. While AI can detect patterns through algorithm, it is not in itself capable of abstract thought. It does not possess a moral compass. 3. If this becomes mainstream, it could generate a wave of misinformation, as people are going to care more about money and kicks than truth and responsibility. One possibility is that artists and journalists end up going to lawmakers and bringing this up, pointing out how it has cost them their jobs. Politicians see an opportunity to gain support from them, and this becomes a political topic as society seeks to regulate AI-generated content.
Fully agree with this. I think we could go one step further and put tags on AI vs Human generated art. Maybe meta content of an image that tells us who created it, and an api for websites to use so whenever you come across an image you can tell if it's an AI or human picture
I mean having a tool for online tests, that compares results and AI with a certain % of matcheing words and if its above a threshold use either a human or another test-layer to compare overall structure of text segments to find out if they have just been rearranged. But ye would take someone with basic NLP knowledge to implement, guess schools wont do that.
“It's interesting to think about how technology like GPT-3 could potentially impact the role of game journalists in the future. While it's unlikely that it will completely replace the need for human writers, it's possible that it could be used to automate certain tasks or provide additional support for creating content. Ultimately, it will be up to the game journalism industry to adapt and find ways to incorporate these new tools in a way that benefits both the writers and their readers” -chat gpt, created using the prompt “Create a comment for a UA-cam video about how chat gpt could replace game journalist jobs one day.”
Journalists are just one field and game journalists are one subsection, the tip of the iceberg. The AI will completely replace the need for human writers, all human writers in all contexts. However just because they are not needed doesn't mean that they won't exist. People will write and create because that process is needed by that individual and offers more to them as a human experience, just not because they are necessary.
The crazy part for me is I already feel like Kotaku and IGN articles are AI generated, I always feel like there's no way an educated human with a writing degree wrote this trash.
There is also the issue of copyright, where AI programs are fed data/art that it wasnt allowed to integrate into its program (including private information as well) Not to mention the people who write the laws arent in it enough to actually see and regulate it currently with the speed of technological advancement combined with the slow process of passing new laws
As a college student who is already struggling with inflation, insanely high house prices and a potential oncoming recession, ChatGPT is terrifying. It doesn't directly affect my field of employment yet, but with how fast this advanced, it feels like only a matter of time till its developed enough to make a large portion of jobs obsolete. Art and Journalism are niche fields which impact a very small portion of the workforce, but what about when its ability to code exceeds the average programmer, it can generate on the spot business reports, it can fulfil most clerk/secretary duties, etc.
This has always been the same in human history. Fields of employment are automated by machines, people cry out loud and then new jobs are created. People start working in these new sectors like in this case for example currating an AI with information or giving the AI the right prompts to finetune the output to your liking. When AI achieves true human level IQ in all fields, any human job will be obsolete and we can all chill out.
I'm honestly surprised AI hasn't replaced secretary duties yet. But I feel like the only jobs that are safe are jobs that require "intense" human interaction like teachers, and ua...honestly idk actors and news anchors already read off a script and can 100% be a pretty face with an AI script. I can see deepfakes getting so good that they can replace actors (though not cost-effective)
Art is not niche at all. All the 2D artists, game designer, 3D designers, animators, even photographs, book writer, scenarists, composer, musicians, sound engineers and so many more will disappear because of AI
When this sorta automation came to factories and was making many blue collar jobs obsolete, they were told to adapt. The same thing is gonna be true for artists and writers in the future, including me. A lot of 3D art is already generated.
The problem with AI art is NOT them stealing our jobs but our artworks!! All day long all I see is one artist after another melting down on socials cause some ass stole all of their drawings and uploaded them to some random AI for people to generate hundreds of drawings in their style for free or worse if their pay for the AI itself! It's absolutely DISGUSTING
Yeah, and the law isn't going to be catching up with AI anytime soon. The U.S. government can barely handle new media like UA-cam and Twitter and Facebook, how will the dinosaurs in the white house handle something this advanced and sophisticated?
The big text generator AI and the code generator AI also both run off of stolen text, so these issues are insanely similar. It's all around a gross situation.
But why though? It cannot even invent new things, all it does is stitch some old stuff together. There is so many thing this AI cannot and won't ever be able to do. For example we will never have good gaming journalism, since the artificial one is based on the real one. Bad input, bad output. Simple as that.
@@Depressed_Spider I think you're missing some points. Yes, as of now AI isn't not far enough to do things you mentioned but you must realize that AI era is at its infancy, and as it grows and develops at an exponential rate it will be able to replace humans in pretty much every field.
@@Depressed_Spider do you think corporations and capitalism cares about new things? They don't and thus creative jobs will be replaced with AI that can do mundane things well enough people already buy broken things like pre orders and cheap shit so new things can just be an old thing with a different coat of paint and people will be satisfied
Thank you for mentioning the AI art part, but I wished you'd also have mentioned the fact that the AI produces these works by generating generalized information from thousands of other human artists who did not consent to their work being used to make a tool like this. That is the other major concern here. These are general AIs, they can't make art from nothing like a human can, they are producing work that is basically the congealed and warped versions of many other artist's works.
Is that also not what a human does? Are every artist's works not simply just a reflection of the their own experiences and their viewings of all of humanity's creations. Humans cannot create without stimulus or without learning, or perception
@@charlieward5439 It is not. Before explaining that, allow me to ask this: If the way an AI uses references is comparable to the way a human artist does, why would we ever strip the human of the enjoyment of making art and give it to an AI? This isn't a monotonous retail job devoid of enjoyment, it's something people do because they love it. This enjoyment can even work two ways: someone comissioning an artist gets joy from receiving the art they commissioned and the artist gets to indulge in their passion while also receiving compensation for their efforts. This is assumming these processes are comparable, but the reality is they are not by means of how they collect references and how they are used. Starting in reverse order; when looking at references, artists and people in general do not naturally have the ability is create complete, identiacal copies of these references on a whim. Not matter how many works by an artist you use, even if they are right infront of you, you will not be able to perfectly recreate their work. On the off chance you CAN do this, then you already have the experience to be an artist in your own right and you have achieved this through your own toil and trials. If you fed an AI every work by an artist, let's say Da Vinci, and then asked it to recreate the Last Supper, it could make a perfect recreation. An AI can recreate any style, so long as it has been trained on it enough. A beginner artist by contrast, will need to toil for long before they can even come close to mimicing the works of others. If they can recreate Michelangelo well enough it becomes hard to tell the difference between a fake and the original, they have have shown their skill despite the illegality of their actions. A human artist collects references generally has more focus, looking for references for specific things that will be present in the final works. Someone drawing a sword isn't going to need references of a wheel UNLESS a wheel is going to be in the drawing. Artists also actively seek out these references on their own, no one is telling them "These are pictures of what a sword is, use them as references", as they know what they are looking for. When an AI references something, it is pulling from a massive collection of images and their respetive descriptions that it has been trained on. The AI itself doesn't understand how it knows a sword from a wheel, it just does this, and cannot easily be trained on new data without being retrained from the ground up. If an AI can tell a sword from a wheel but you want it to draw a shield, the AI must be retrained from the ground up on a modified or new dataset with images and descriptions of swords, wheels AND shields. Imagine if a human who wanted to draw a knight had to relearn the look, use and purpose of everything associated with a knight (armour, weapons, AND what a knight is) just to draw one. On that subject, these datasets are often massive, dwarfing the number of references an artist would use for all the works they made in their entire life. The amount of data they have can number in the BILLIONS. Laion, a dataset collection Stable Diffusion was trained on (just one of them mind you), has Laion-5B which is around 5 billion image links and associated descriptions. The issueis everything and anything can find there way into these datasets because the categorization is indiscriminate. Copywrited works and even PRIVATE MEDICAL IMAGES have been found in 5B. While the latter is bad enough, those copywrited works are where it gets concerning because these are works that a normal person could not copy to a personal blog without possible legal risks. This is were I would like to make a point VERY clear: The issue is not the AIs but the dataset they are trained on and the people behind both. Laion is able to get away with using copywrited works because their status as a German non-profit company, which affords them certain legal exemptions that other for-profit companies would face. This data, derived from another non-profit company called Common crawl, is now being used IN for-profit projects like Stable Diffusion whose creator, Stability AI, funded a large part of Laion. They are using a complex web of companies to skirt around the law, abusing these legal exemptions for their own gain. Worse is, some AIs have not release any information about which datasets they've have used, meaning we have no idea what public or private contents these AIs have been exposed to. Human artists and AIs do not operate the same, as the AIs have access to stuff normal people would not and use them in ways that would lead a normal person to face legal ramifications for. These AIs, thorough the use of loopholes, make use of non-profit privledges to gather date to be funneled into projects by for-protif companies while those non-profit companies are being supported by the same for-profit companies using their data. Edit: spelling corrections
@@pipes2051 you literally didn't answer the question, explained AI badly enough that I feel like you don't really know much about how learning algorithms work, and posed an entirely unrelated question. AI learn by viewing datasets and then finding and creating associations between different variables which describe the datasets (the more times the association is made, the stronger the association). This is not entirely dissimilar to how the brain recognises patterns and learns. It then uses this to create whatever people ask it for based on prompts. Why is an AI learning off of intellectual property so wrong? It does not copy the property, it only views it. An artist does not have full legal control over what anyone can do with their intellectual property and their IP's likeness, only over other people either making money off of their IP, people pirating it, or people distributing it without their consent, depriving them of income. This is how it should be, no one should have legal control over what a private individual wishes to say, draw, or write. If I choose to train an AI off of other people's intellectual property as a private citizen, attempting to use legal means to stop me would be a VAST overreach of the law on individual freedom.
@@charlieward5439 The issue with an AI using copyrighted works for training is that they are not private individuals who are exempt from legal risk (so long as they are not used in a way the brings). They are the products of for-profit companies. These companies have no legal right to make use of copyrighted material, especially those by other companies, or images of private information in their products. They get around this by using data gathered by, and in many cases providing funding for, non-profit companies who compile the collections used to train their AI. They have access to materials the average person would not. If an artist used images of medical documentation, especially those pertaining to specific people, they would bringing legal action upon themselves if ever found out. What legal action would this imply? I don't know; part because I am not a legal expert in any measure and part because it depends on location. As for how AI actually learn, it seems similar on the surface. But this hinges on AI and humans experiencing stimuli in ways that are identical or at least very close. If a human were to see an image of, lets say a spider, they could still recognize a spider if they saw one in a video or in real life. If they were to see something that looks simliar to a spider in some way, like an ant or a scorpion, they would recognize it is not a spider. AI cannot do the same thing. An AI trained to recognize spiders using images can only recognize spiders within images. If you were to show it a video of a spider, it would not be able to recognize the spiders in it. Additionally, show it an image of an ant or scorpion, it will be confident it is a spider. If you wanted the AI to recognize the differences between spiders, ants and scorpions, you would need to retrain the AI with images and descriptions for all three. Same goes for recognizing both images and videos. Also, a person's perception of an object or concept can be changed. If a person who has positive or neutral feelings on spiders is bitten by one, that experience goes on to change their perception of spiders in a (usually) negative way. AIs do not have perceptions of objects and concepts the way humans do nor can be changed by experience. If an AI associates spiders with negative feelings, its because it was told that by someone who has or understands those associations. This is because, at their base, humans are only good at pattern recognition and learning/adaptation while AI (and computers in general) are only good at data storage and arithemtic. When an human artist creates a drawing, they are pulling on experiences and images they have obtained over their life. New experiences can easily be added to a person's mind and old experinces can coloured by new experinces, alongside the feelings associated with the experiences. When an AI is asked to make an image, it does so by drawing on data it was taught to recognize. If it is asked to do something outside what it has been taught, it will not be able to do so without being retrained on data that includes the new addition. It can certainly TRY to do so, but the results won't be akin to what anyone would expect from a human artist. Belive me when I say this: I do not have much issue with the existence of these AIs. However, I have seen too many over hyper advocates using this same argument to put down artists concerned with both their futures and how these companies are using their art.
@@pipes2051 you are actually just wrong about how AI create from prompts, they do not store any of the data that they are trained on, they simply learn from the data and are then able to create requests from what they have learned from the data. They do not store the data from the dataset. You seem to think too, that humans are born with innate ability which they are not, if you were to isolate a human from external stimulus whilst they were young and teach them only to recognise spiders, then show him an ant (which he'd never seen before), you would also find, funnily enough, he would also probably say it was a spider. You seem to view learning algorithms in the same way you view standard operations in a computer, but it's really not that simple. If a human ands up with a phobia of spiders from being bitten, this is most likely because of the association the brain has made between recognising a spider and the pain from being bitten, this is an unconscious association and the person with said phobia does not recognise why they have the phobia. You can actually do a similar thing with learning algorithms, if you take a learning algorithm which has been built to navigate a play space with a sprite and accomplish a goal, but has a health value which makes it fail it's goal if the health value hits zero and put spiders in it's play space as well, it will not have any positive, nor negative reaction to the spiders at all. However, if you suddenly change the rules of the play space so that the algorithm's sprite can be damaged by the spiders, you will see it learn very quickly that spiders should be avoided and it will avoid them at all costs, same as a human who has been bitten. Interestingly, both the human and the Algorithm, unless explicitly instructed to do so will each avoid any spiders aggressively, even if the spiders are no longer able to inflict damage or will not bite as both has made the association between the spiders and pain/damage. The only difference in these two cases is that a human would say that the human's reaction is emotional whereas the machine's is not, however, it is an interesting differentiation to make as the basic premise is practically identical, the only real difference is the relative complexity and disparity in imperfection between the brain and a learning algorithm
I'm no expert AI engineer, but I'd like to think I'm well educated as a software/computer scientist & engineer. Being so close to the field in my career, I can safely say that humans will most likely NOT be replaced by AI anytime soon. For those that aren't entirely aware of how AI is created, simply put: AI is trained on a very large set of data and it finds patterns within these data sets and with human created algorithms/techniques and math it is able to use these patterns to recreate things with some level of precision. The nuance issue that stops AI from taking over humans is that they need human generated algorithms and data. The AI article and art Charlie is talking about required most likely thousands if not hundreds of thousands of sample data to generate something. This data acquisition can be both illegal and unethical (illegally violating IP laws and unethically because you are appropriating someone's work without permission). AI being used legally to replace humans would first have to acquire these datasets legally and hopefully ethically. Bad actors will always exist so this won't stop someone training AI using illegitimate means, but those AI packages would likely not be accepted by the entire world (I could imagine the US would not allow this, but other country governments might). Lastly AI is highly specialized, meaning that the model is trained to one thing, and ONLY one thing. Each AI has to be trained on unique datasets for their highly specific task, so even the argument of "AI that creates AI" is a bit far fetched at this point in time given that lack of data to train them. I foresee a future where AI does not replace humans, but becomes integrated with everything we do (Kinda like cyborgs but mainly with highly specific software for certain tasks). TLDR: AI right now has ethical and legal issues to overcome first. AI has come so far in a short amount of time because humans are awesome and supplied them (knowingly or unknowingly, again ethical issue) with loads of quality data, and I believe AI is generally our next evolution not our replacement.
@@sloth1667 I'm not entirely sure the legality of posted online content, you'd probably want to discuss that with someone that works with IP laws (I ain't no lawyer). Ultimately the website probably has some clause saying that the work can be used in "fair use" projects, but "fair use" was first written under the impression that a person was directly working with the source (not a person programming a computer to use the source). There are loads of ways that content is protected online, and some allow for legal use for AI training I'd imagine, but the larger issues at hand is when AI is trained on data that falls into the grey areas of IP law. You may be right in certain examples for AI being trained with posted online data, but you also have to consider the ethical side as well. Something can be legal, but unethical. I have seen AI do things that seem borderline illegal also, like recreating art in the style of "Anato Finnstark", which could imply that the model was trained on most likely copywrite images from Wizards of the Coast and Dark Horse. This is a cherry-picked example mainly to highlight a potential legal issue. These neural networks aren't super new (10 or so years in development), but they have exponentially improved recently so all the laws and standards haven't had time to catch up.
@@full-timepog6844 they have a base reason to assimilate/kill everything being to control until it is stopped which most A.I are directed todo in some way right now if it cant alter itself then it might create another and self terminate or have this lesser A.I to fix the flaw, if it is to create utopia then it might enslave/assimilate then release or raise a new generation brainwashed to be able to live with the world it created based on its interpretation of utopia and there is little media for it to come to a conclusion we would want and a lot of media to corrupt it that we use to entertain ourselves
Lol I was just thinking about this. It turns out that if your job involves inputting things into a computer, or applying reason, or knowledge, AI has already shown the ability to replace that. People who don't think it can have not talked to these new large language models. It still seems bizarre but laborers are protected by the fact that somebody still has to build complex mechanical devices to do any given task in the physical world. Turns out all it takes to replace computer workers is to deploy a new software package. We're in for a wild ride lads. One of those 'if I didn't laugh about it I'd be crying' moments.
@@noname-gp6hk >It still seems bizarre but laborers are protected by the fact that somebody still has to build complex mechanical devices to do any given task in the physical world. It's even more bizzare considering that most of those laborers are one bad immigration policy away from being yeeted out of job market, but they STILL act smug about it.
We should train Ai to be able to detect things made by Ai
That AI will train AI to become undetectable.
There is a whole AI subfield revolving around this idea: Generative Adversarial Networks (GAN's). It's actually the general architecture behind how deepfakes came to be
Why would laymen care? They'd just appeal to a "good enough" product.
“If it's too good to be true, it probably is” ❌
"If it's too good to be true, it's probably made by an AI"✔
@@neo-didact9285 Not sure why some people fail to understand that a lot of laymen actually care about the people making products. People wouldn't care about AI art if they didn't care about artists.
The craziest thing about it for me is that everyone always thought that creative jobs would never be touched by AI, but it's legit the first thing it went into.
I can see how useful it is for art and writing but it is still a tool for now, and should be treated as such. Otherwise it's like presenting a chisel to a client who's asking for a wood carving lol
Because I guess they want to screw with creatives.
I oddly am convinced some just hate human creativity.
@@crowdemon_archives
It's a tool with questionable ethics and potency for exploitation while devaluing.
@@crowdemon_archives It's only useful to people who aren't an artist that's why some people defend it to the end and try to justify the stealing of artworks from artists without their consent to feed the AI
Most artist just use AI for fun and inspiration but they would never replace it with making art
@@crowdemon_archives AI for art is like auto tune for singers. This isn’t going to be treated like a tool, it’s going to be like cgi in big movies, overdone and soulless.
A few years ago: "Journalists, learn to code."
Now: "Journalists, learn to be code."
Code, learn to journal
@@CircusofPython journal, code to learn
@@mihael64to, journal code learn
more like: code learn journalist
I am become code
Imagine going through college today and just having AI write all of your essays
GPT can't write essays. At least not original ones. You'd get dinged for plagiarism and out yourself the moment the professor asks any questions about your process. Not to mention citing sources.
Imagine being in school and using GOOGLE. 🤯
@@chatboss000 It outputs text based on what it learns. It doesn't just paste things from the data it was trained on. I got my very last college essay during the summer written by GPT-3 and got a 100
@@user-rf4vc7mt4d 0.0 oh
@@chatboss000 just use a paraphrasing tool
AI replicating a job based off human emotion is scary
Game journalism has no emotion
Art not game journalism bruh
AI being able to mimic and create various original art styles is the gateway towards becoming sentient robots
i mean social media literally is ai manipulation to keep humans on their sites for as long as possible, why are people surprised about this stuff now lol. Its just before you had to build your own AI, this is just the first time the public has gotten access to what comp engineers have been working on at google etc for ages.
Then nerds will have nothing to complain about when a game they like doesn't get a good review
Being called a "bot" is slowly becoming a complement
Lol
Lol
FYI: There was a time when being accused of using an aimBOT, without using one, was a compliment. (Back when CS:S was the main shooter)
With AI the field to apply computers to just got broader.
my prediction is that all competitive gaming will just be different AI players. We will give them all names and humans can only bet on the matches and watch. I'm not talking some cheap CS:GO bot either.
@@TheBigLou13 Its still a compliment
There's something just really depressing about how in the beginning we thought AI would free up people from manual labor in order to pursue more creative and artistic endeavors but instead we now live in the world where all those endeavors are being automated now by AI while for some reason the world is very insistent that the most depressing, grueling, and even dangerous jobs *must* still be done by humans.
Maybe in the distant future, there will be physical moving robots with ai designed for specific tasks. Ai is amazing but it has its limits as long as it’s bound to a computer
@@legendarynick2798 Well said.
Right? I thought AI would free me from having to do the things in life most people don’t wanna do. Creativity is something humans always have done and enjoy doing regardless if there’s anything in it for us. Why did it have to be that? It’s exactly the opposite of what a lot of us expected and it even feels like a future with AI isn’t what we thought it would be
Don't blame AI. Blame the people who prefer to make people do slave labour instead of using AI that can do the same thing, only at a marginally higher cost.
@@user-xn3rs7tu5w blame greedy business owners. 30 cents and slave labour is a lot cheaper than 2 dollars and an AI
Man artists just can't catch a break. First the art thieves, then the NFTs and now AI art. As someone who likes fanart it's been painful to see my favorite artists struggle with so much shit.
True but it’s also true that many artists took advantage of NFTs and made money selling their art as them. Many artists suffered from NFTs but it only makes sense that some followed the trend and made good money.
fr. Artist can't catch a break.
I mean, at least this AI can’t generate fanfiction lol.
@@laexdream It most certainly can, if the AI's training data contained knowledge of let's say "Lord of the Rings" and you prompt it to write a short story about it, the AI should be able to do that. ChatGPT can do a lot of things. Muda made a video showing of that ChatGPT can give you advice on breaking into a house. (don't try this at home)
Oh and there is a Video that shows ChatGPT generating code for a Minecraft fly hack on YT too.
Aristry will always be around no doubt
I really want Charlie to have the ai make a script for his video, and for him to make it and publish it, and not tell us for like a whole month to see if we can pick up on it
Someone made an ai deep fake video of charlie a few months ago and it was scarily similar to the real thing
mkbhd did that. check out his latest video.
i tried making a script for him, but ChatGPT doesnt know who penguinz0 is :(
@@lux0rd01 We know those videos use voice cloner AI, but the scipt is more likely written by a human, though it wouldn't be hard to feed an AI lots of transcripts of charlie talking then have it spew out his signature verbose lengthy insults.
@@MumblingSolipsist I was referring to one that was both a deep fake AND an ai written script
We’re really starting to enter the AI era. Its insane.
*Ok_Dont_Read_My_Names!!.*
.......
Jesus Christ loves you
@PizzaByte you get no play zero hoes zero bitches
It's ironically hysterical that this comment attracted a lot of bots.
@@underwatermonkey3443 You're replying to someone who will more than likely never respond or see that... It's called a bot for a reason. Report them maybe someday UA-cam will decide to actually do something about it for once.
People could so easily write argumentative essays for school with this and it wouldn’t even be flagged as plagiarism because it’s actual original text. This is insane
Ai combines text from other sources it would be possible to reverse engineer it most likely
@@juckya9660 Highly unlikely. You have to keep in mind that the neural network isn't simply copying text from other sources and compiling it into a giant paragraph. It's taking certain prompts from many different sources and generating new sentences from it. It's basically impossible to "reverse engineer" because it's a brand new sentence. As it currently is, plagiarism checkers are only effective when it comes to large chunks of text. The smaller the components, the harder it is to identify.
Some day, schools are going to have to warn students not to use AI to write their essay papers the same way they warn us now about plagiarizing. Except it’ll be harder to pinpoint if the writing is “plagiarized” from the AI, since it spits out newly generated text all the time...
I can see this becoming a more prevalent issue in the future, and I know, because I’ve used it myself to get through highschool. Hell, I let AI write my graduation speech as valedictorian, and nobody found out.
Companies hosting these services might have to keep track of what’s being generated or have the text saved somewhere the public can see. But until then, you can pretty much take advantage of this resource with almost no repercussions. Nobody will know.
Oh some college students were already using it to do their homework. Then realized it was so fast they could do everyones homework, so they made it a business until they got caught and stopped by the college.
I highly doubt u used AI to form your ‘valedictorian’ speech lmao
Some day, AI will be in charge of our schools and humans won't be allowed to write stuff because we're too slow and dumb.
They'll focus harder on testing in college, give them time.
@@shmeatr2463 why do you doubt that? generative nlp models aren't new
A few days ago, I used this AI in an online Dungeons and Dragons group with a close group of friends I've known for about 5 years. I gave the minimum backstory of my character, world, NPCs, and other player characters.
I started the game like any other, though once the session was in full swing, I kept copying and pasting everything that was happening in the session, into the AI's prompt, to keep it well informed on the status of the game.
Anytime my character would make an action within the game, I would type "Provide me an appropriate response to X situation, with as much detail as possible."
Using the minimum information of the game, the AI appeared to grasp the entirety of this fantasy world we built over the last 5 years, and made appropriate actions with the POV of my character in mind.
In fact, with very minor adjustments and a few 'Try Again's to keep the AI on track, it was capable of getting the party out of a very dangerous situation, while keeping to the morals and ethics my character upholds. AND ADDITIONALLY, my friends could not tell the difference at all, and has no clue it was an artificial intelligence playing in my place.
We were in shock of what this AI can do. Keep in mind that there is a much more advanced version of this AI that has yet to be released to the public. Frankly, horrifying, but strangely beautiful.
What AI is this? As a dnd player u got my interest
@@joaoborges1764 it's called chatgpt its free online just need to make an account
At least say the name?
@@jonathanmastergame lmao just look at the replys
Been going on for years. Social media is 90% fake comments.
There's so much footage of Charlie sitting in his chair and talking, I'm sure AI could compile all his footage and sounds and take over this channel. It may already be happening.
Didn't he already do a video on that?
It already has
@@flameshana9 no that was deepfake I think
@@volty3454 Yea not sure why Char;ie framed this video this way, does he have something against journalists? How long will it be til the "Streamers Will Be Replaced by AI" video drops? Will Charlie still say its a "really good idea" then?
I mean did you notice how positive to Ai this video is? He's already been taken.
as an artist i'm sad. as a programmer i'm happy we've come so far. weird world we live in.
As a programmer you should also be sad because chatgpt is gonna take jobs of millions of junior software developers
@@fallen2624 it's not going to. if you looked into it, it can replicate very simple patterns i has already seen before. It cannot even solve the newer leetcode Easy problems. It's really weak when it comes to extrapolation. It will be used to automate things, but it will not replace programmer jobs, just make them easier.
@@junkoe3808 you should take into account how fast it grows and develops. Sure, right now it's just a tool that makes work easier and faster, but even this is enough for demand for developers to decrease, especially for those who are just starting out. Since 99% of the work that juniors do can already be automated using this tool. Imagine what's gonna happen in 10-15 years?
@@junkoe3808 Haha, many thought the same thing about art. Just u wait. Programmers will become their own Ai's bitch. At least junior level programmers.
@@junkoe3808 you are overlooking the bigger picture by focusing on now and here points. Literally strawmanning out of your own worries so you don't have to think about them. Worst case scenario AI will relegate most humanity to physical labor under big tech if this trend continues uncontested, as for the best case scenario - every blue collar job will now be contested by an AI that forces a rather brutal entry bar.
I can't stop laughing at how the AI jumped to the conclusion that planet Zebes is a stereotypical depiction of East Asia.
Omg fr
Ign journalists *nervous sweating* (i saw an article about video game racism similar to this on ign)
"Flesh is cringe, machine is the future"
Such a powerful quote
From the moment I understood the weakness of my flesh, it disgusted me. I craved the strength and certainty of steel. I aspired to the purity of the blessed machine.
Your kind cling to your flesh as if it will not decay and fail you. One day the crude biomass you call a temple will wither and you will beg my kind to save you.
But I am already saved. For the Machine is Immortal.
Tech priest moment
@@film57r7 you zoomer techprist have no understanding of what it means to be a true magos back in my day we had to lobotimice the serivtors ourself
PRAISE THE MACHINE SPIRIT. I HOPE IT REPLACES ME SO I CAN BECOME A NEET
Never upset the machine spirit. Respect its will, and it might just respect yours.
As a kid i used to be scared of AI, now i'm terrified
@@Amneso just wait until it or we create a whole new computer architecture that allows computers to do that
Meh, grow up! And get a high caliber gun.
@@Amneso you can't even fucking define what being sentient is. don't act like you know shit
@@Amneso This is exactly what they were saying in Terminator 💀
@@ireallyreallyreallylikethisimg It either goes terminator or bicentenial man.
as someone currently in college working on their journalism degree, this isn’t making it any easier for me to not drop out :,)
edit: I’m going to reiterate what I said in a reply below. I appreciate it and I know some of you come from good places (granted not all of you) but I want to make it clear I will NOT be taking advice from strangers on the internet. I made a lighthearted comment because the video coincidentally pertained to my major. College is hard for anyone, I’m not actually dropping out. I’m going to make my own decisions on my life and be happy with them no matter what you say here. Please respect that. Thank you!
Unless you getting killed by cai/fbi it's not worth it.
As an undergraduate researcher in AI, I am sorry
But also as an undergraduate researcher in AI, I am very glad this is what I chose lol
Don't worry about dropping out, you should be concerned about taking this opportunity to stand at the forefront of this technological advancement and make the most out of it. Make AI your tool and not your replacement. Do your research and be educated about how AI is (and isn't) affecting your field
You can always bring back investigative journalism and bring the truth to people who are too stupid to appreciate it. You have until AI develop drone bodies and a sense of vigilante justice at least.
good journalism can never be replaced.
AI generated videos could easily become big on youtube too in the future, to be honest im kind of scared how good that article generation is already. Imagine how easy it would be to convert that to speech and generate a character that says it. AI moist?
There's already a video of an ai generated moist critikal video. It's pretty freaky
There's already an Ai generated kids gaming channel called Bloo. Quebblekopp made it and its got a bunch of subscribers but the views are kinda eh. Videos are terrible but it's on par with other shitty mass produces child friendly gaming UA-camrs
I can write an essay about how AI generated videos could easily become big on youtube in the future
AI generated videos have the potential to become very popular on UA-cam in the future. This is because AI technology has advanced to the point where it can create videos that are almost indistinguishable from those made by humans.
One of the biggest advantages of AI generated videos is that they can be produced quickly and at a low cost. This means that creators can churn out a large volume of content in a short amount of time, which is essential for success on UA-cam.
Another advantage of AI generated videos is that they can be highly personalized. For example, an AI system could analyze a viewer's preferences and create a video that is tailored specifically to their interests. This level of personalization is not possible with human-made videos and could be a major draw for viewers.
Additionally, AI generated videos can be made in a variety of styles and formats. This means that creators can experiment with different types of content and see what resonates with their audience.
However, there are also some potential drawbacks to AI generated videos. For example, some viewers may find the lack of human involvement to be off-putting. Additionally, there are concerns about the potential for AI generated videos to be used for nefarious purposes, such as spreading misinformation or propaganda.
Overall, AI generated videos have the potential to become very popular on UA-cam in the future. They offer creators the ability to produce a large volume of personalized content quickly and at a low cost. While there are some potential drawbacks, the benefits of AI generated videos are likely to outweigh any concerns.
@@theclassyxenomorph1301 That video was fake, Charlie got it wrong. It wasn't as seamless as the video he showed. Ai can edit videos and they can create animations but It's going to take a longer time that our lives for AI to be able to take someone's face and make it appears as if they're making a video and speaking lol and doing things lol. Who knows, maybe Metaverse will figure that out by then LMFAO 💀💀
This is probably testament to how robotic game journalist articles are rather than how impressive AI can be in generating a random wall of text
I don't think so. Its just really good at imitating the articles it learned from.
It's both.
You probably haven't seen what these new large language models are generating. Random wall of text is not an accurate description.
AI can pretty effectively synthesize takes based on standard level thinking. If you're not particularly creative or insightful you can be automated pretty easily, and even if you are you'll use AI to assist your work. I haven't seen indications of AI coming to independent conclusions, but it can effectively synthesize and sum up what's fed into it.
The Robots are just machines. They can't be bargain with, they can't be reason with, they don't feel pity or remorse and it abouslety will not stop ever until someone tries to detect weather if sentences are robotic or human.
I wouldn't be surprised if Charlie was an AI generated program this entire time.
But isn't he Jesus?
@@steelheart538 robo Jesus
Well he's already talked about the A.i videos made of him
I legit saw a generated moist meter video the other day and it was pretty accurate
That'd definitely explain the voice lol
In art field it s even more annoying frustrating baffling because the ai is trained based on other artists
And not only dead ones.
Artists who could be struggling even more to find a job that ai TRAINED BY THEIR ART stole from them.
Artists have their hard work taken w/out permission to create app/webside to replace them.
If AI does it better, then who cares find a new job
@@Tylerrrrrb if the only reason "ai does it better" (doubt) is because works of artists that didn t consent to having their pieces used then there is a problem. Lot s of ai bros straight up put artists names in prompts to copy what they mastered. And now want to profit off it? If you don t see what s wrong w it then man pls work on critical thinking skills
And ye im not in art field bc i fear stuff like this. Instead im stuck in med sql💀
@@Tylerrrrrb imagine the career that you have honed for years with little actual economic incentive suddenly getting obsolete because a guy trained a bot to copy exactly how specifically YOU do it, and then proceeds to churn out a ton of money out of a skill you developed.
@@vrygon8893 maybe don’t hone in on a Career with little economic insensitive for years? Obviously ? If people did that , then they’re morons tbh
@@munmun__ stop with this copyright argument, u artists also "steal" from other artists work, dont come here pretending u dont use other art as reference for inspiration, the only problem u have is that the machine is doin it faster than u can ever do it, if u are goin to say AI steals from artists without consent, then all artists are thieves aswell.
I write articles as part of my job, and my manager told me last week that he’s testing an AI that writes these pages much faster than we do. The prospect of losing a huge part of my job to a fricken robot is kinda scary.
The scarier thing would be if the ai had access to social media data and was able to tailor make articles for each person depending on their preferences.
You could do this with current technology. Scrape a person’s social media profile from the internet and ask them survey questions, prepare an RSS feed catered for their tastes, and use ChatGPT prompt engineering to reword the RSS feed to be written specifically for you. It’s a brilliant idea lol
@@williamuemura7644 This actually sounds like a talk Tom Scott gave about the future like 10 years ago. Scary, but also kinda cool.
@@impishlyit9780 Fortunately, these AI aren't capable of critical thinking. They're only able to recognize patterns and copy them.
That means that AI depends on human content creators for reference material. In the future, AI will have to train off of AI-generated garbage, computing costs will go through the roof, and our AI overlords will collapse and die.
OpenAI has historically spent one year between releasing their GPT models. They've spent over two years on developing GPT4, and results aren't very promising. I think that the AI boom is officially behind us, and that progress will be permanently slower. Game journalists are fucked tho
I think this is kind of what facebook is betting on. I think they're betting on technology being able to read your feedback to content, react, and create personalized VR experiences unique to each person. Imagine the dopamine hit to be able to slip into utopia every time you put your facebook helmet on. The greatest experience your brain can possibly imagine. You'd never want to leave. It *still* sounds like scifi BS, even to me, even now, but it's pretty clear that's the next step. Everybody makes fun of 'the metaverse', but think about where this is headed. And that's probably the scariest thing I can imagine. We're a couple decades from losing our humanity.
@@williamuemura7644 why are you so confident, I'd love to know your insight?
I understand AI at the moment is just copying and regurgitating information fed to it but isn't that most of human work already? Also what if future AI is taught how to critically think, then it's potential is massive
the scary thing isn't how strong AI is, it's how easy it is to misuse it, like winning art competitions, or political impersonation, I still want to be optimistic and say we'll learn something from this, but it really seems like we're barreling towards another crisis here
What was the first crisis this is following?
Exactly, and apparently they’re already using it to make deepfakes of CP :/
@@scatdawg1 I'm just talking about crisis in general, just seems like there's a lot these days, the pandemic, ukraine, just to list the big two
Im going to misuse it to farm gaming articles
I think the best way to fix it is to make it illegal to use data from other people’s content that you don’t own, and treat it under copyright laws as if it was just plagiarism. If you only use your own artwork for an AI to generate off of, fine, but if you use other people’s for the AI to base its art off of, it’s pretty close to stealing their art
For a second there I almost started to get mad that Kotaku or whatever site would write something so ridiculous and I had to remind myself that it was AI generated.
The counter argument would be, why would people tune in to articles written by in this case an AI that has no actual understanding of the topic at hand? Why would the opinion of something that doesn’t have opinions matter? Or is the idea that they would associate a “fake” journalist to an AI written article?
because it is possible that an AI will have an actual understanding of the topic
because an AI is unbiased ?
@@luxberry9634 I don’t know if I would describe current AI as unbiased.. I don’t think AI has a “bias”, it’s code a human put it to mimic in this case human journalism. Even if it’s perfect it’s still not a real opinion.
@@luxberry9634 Dude did you watch this video? You can already have an AI echo the political biases of a journalist pretty convincingly. And who wouldn't hire AI to write glowing articles about their games so they don't have to bribe a human yes-man? Obvious profit incentive.
Game journalists are bottom o' the barrel writers and replacing them is not hard.
The fact that AI was able to accurately replicate the average videogame review by attacking the game with absurd accusations of discrimination and somehow put the blame on a public figure while leaking their biases and beliefs into the material is kinda scary. It knows *exactly* how to behave and what to say. Either humans have become that predictable or AI has become too good at understanding us while we understand little of it.
Humans are predictable, we're the ones with the routines. The ai is only imitating our routines, we know why it's taking the action too. We only haven't made the bridge to understanding how does it make the association between what you ask and what's the response, we just say if it's the right answer or not. We're teaching it to be the most generic human, there's not a shred of intelligence there
All it does is look at articles with those keywords, process the words, and try to use said words or phrases while following language rules it has learned or been trained to learn. That's why it's such a generic piece. If anything I just think it highlights how bad journalism can be. If the quality of the content people are writing is a joke, ai will treat it as a joke too
@@knightrider9876 The real answer is that the roles we design for ourselves are simple enough for AI to fulfill. In other words, people and AI are both intelligences submitting to a reduced role for the mind, when the AI does it just as well or even better, it just means the AI has less to sacrifice to become the role, than we do. AI lack social identity and other mechanisms (for now), they have so to speak less to worry about and therefore less reason to hold back when "specializing" to a certain category of tasks. Until then we're merely comparing the ability for the nerves of our back to facilitate reflexes compared to voluntary movement - of course the nerves in the back are better at facilitating reflexes than the mind.
People think their hands stupid, but there's intelligence involved that rivals that of our strongest AI when it comes to our hands. Same with our vision and ocular nerve, the ocular nerve doing far more intricate visual processing than any AI we make today - and that's before the nerve crosses the blood-brain barrier. Our AI aren't that strong yet, to even rival that of body parts.
Well, both mean the same, relatively speaking
id say for something like kotaku, woke is predictable,
Not only does it write well developped essays and articles, it can also write pretty complex code in various languages too. My creative coding professor litteraly warned us NOT to use it for final projects because of how easy it makes it to cheat the plagiarism checker software of our school
do it anyways
@@AustinCameron
It's comforting to know that it still can't actually understand the code it writes.
The real issue is for the next few decades to come: What will we do once it _can?_
So he warned you not to use it because it can't be detected by the plagiarism detector? I feel like that's sending the wrong message...
@@AustinCameron that I didn't know, kind of a relief it doesn't write perfect stuff at all times
@@bellicose4653 I know right?? Though I didn't use it I'm 100% sure some people in the class are going to💀
It's important to address the potential negative impacts of AI on employment and find ways to ensure that workers are not left behind as technology continues to advance. (this comment was created using AI)
I’m honestly terrified if AI comes to music too, how would I prove I made it and didn’t use an AI? Someone without any musical knowledge could possibly use AI make beautiful pieces and get credit that they don’t deserve. I legitimately think that the government has to do something about this eventually or else human expression and art will be meaningless. How am I supposed to be proud of my hard work if people are constantly accusing me of using AI?? It’s truly terrifying
Even if there were international laws against AI, the government can’t stop something that has little to no trace of being artificial
Get a job at Taco Bell instead.
Yeah valid concerns, it’s pretty depressing. I’m a music major and theres an impending doom I feel about it and a sense of hopelessness. It makes human expression feel less special when an infinite amount of art can be created on a whim. 😢 I’m holding on to the hope that people will find merit in human made artistic endeavors even if it can’t match the quantity of ai art. Ai art will never have the heart❤ that human art has, and hopefully that is enough of a motivation and fulfillment to keep being creative.
I've worked on AI automatic text generators in the past. One was an automatic poetry generator that could emulate the writing style of any author, at least in theory. Results showed it was promising, since people couldn't tell the difference between one real author's work and the AI-generated poetry.
The catch? That was 4 and a half years ago. I thought we wouldn't get a system as advanced and convincing as ChatGPT until at least 10 years from now. The pace of AI research has reached breakneck speeds.
The ai is probably helping itself become more advanced lol
I think that it's a snowball effect. The stronger and more intelligent the AI gets, the faster it will develop. There's a theory where computer speeds double every 18 months, it could also ring true for AI
I agree the development of this technology is insane overall.
It will only get faster. And faster. This is accelerating. There is no closing this door. AI is here and I am willing to bet this by summer of next year it will be self aware.
@@fizzle7421 not yet
I'm sure game journalists will be missed dearly
NOT
😂😂
@@yorusaka3554 😂😂
😂
@@emilfrederiksen.1622 😂😂😂😂😂
4:00 The article written by the AI is similar to something you would see from journalist websites trying to bait reactions from people where they want to talk about an issue or complain but they don't really know much about what their talking about only some highlights that they can warp to the point they are trying to make
I got this message from ChatGPT when trying to write an article about a movie. "Since I do not have access to the internet to research the movie, I will have to rely on my own knowledge and imagination." Now im thinking everything over.
Also got this when asking to write an AI message. "Hello! I am a large language model trained by OpenAI. I am not a true AI, but I am designed to assist with a wide range of tasks and answer questions to the best of my ability. Is there anything I can help you with today?"
I used to write copy for an agency and right about my birthday in September I was told I was being laid off because they found an AI copy program that was way cheaper. Skynut is real, folks.
They told you that?
Really?!
Skynut lol. Sorry thats really deep but i couldnt avoid that
@@Therealferm I was referencing his Skynut music vid hehe
@@mitchelkvedar674 yep. Worst part is the AI's copy is pretty good 🙃
"The scientists were so pre-occupied with whether or not they could, they didn't stop to think if they should." - Dr. Ian Malcolm
They should it's just that with the current capitalist economy it will be used for profit and not good. Are rich people gonna care to pay working class people if they can just replace them? The one thing we have over the companies their dependency on our labor. If we are replaceable we lose our voice and since we don't have power we become nothing.
They should go further. Lets see if we can't make skynet by the end of the decade
@Vessel of Amala except it's rich people giving these ai companies money not consumers lol
@Vessel of Amala lmao how good must the bottom of their boots taste to say something like that? The consumer has no power in an economy run by the producer class.
Tell me, how easy is it currently to purchase purely ethically sourced food that isn’t a result of environmentally damaging practices or have been stuffed full of chemicals to keep it sellable for longer? And don’t say “make a garden, buy your own animals.” Cause that is completely unrealistic for someone who is working full time just to survive. If you do not have the purchasing power to compete then you are screwed and your only option is to buy from cheaper, less ethical sources. This is the prevalent systemic issue in the current paradigm. Attempting to consume and “use our purchasing power” elsewhere is impossible.
Its literally the most prevalent trend in capitalism that as technology is iterated upon it is made standard and therefore required. When everywhere has ai because its cheaper what are we supposed to do as consumers? Just starve?
Its just so funny how you can look at how dominant and powerful corporations are and still somehow find a way to blame the individual. They aren’t gonna give you a job for defending them online dude. Get their boot out of your mouth and wake up.
There's worse to come check out the channel StarTalk and listen to what these scientists wanna do to the human body
Okay I'll be honest I've brushed this shit off for a while now and thought that especially artists were over reacting about Ai, but if it can create that entire article off a single prompt, that is genuienly terrifying
It learned from previous Kotaku articles.
I think people are missing the real implication of this. If AI can reproduce game journalist articles then perhaps we need to analyze just how incredibly simple, regurgitated, and rehashed most game journalism is.
AI development is crazy fast now, but if you tell it to reproduce something that had legitimate critical thought and not some buzzword hit piece like most articles are now, you could definitely tell it apart.
@@cosmic_sloth so... if we go back to thinking about artists, does that mean that most art is "incredibly simple, regurgitated, and rehashed"?
like on one hand im not a fan of the sorts of works AI produces but just cause some journalism sucks doesnt make this not a legit concern lol
@Sonny Yeah but it's not just game journalism, this ai can generate 100's of lines of working code simply by describing the program you want it to create, it can summarize text you give it, explain some fairly obscure topics to you, you can even use it to tweak the style and feeling of some text you give it. You should give it a try yourself or at least look at some more examples online before you make such confident statements.
There's plenty that have been doing that for a long time. Jarvis AI (I think i'ts called something else now) was shown to me by a friend and it puts out seamless blog posts for her blog. There are better ones that target creative writing, and I haven't played with them yet but I heard they help writers a lot. It's definitely already been going on for a while now and will only get bigger. None of the writers talking about it on reddit seemed concerned about that though so it must not be that big of a deal yet. They were very clear that it's still more of a tool than a replacement.
I’m a software engineer and I been toying around with chatGPT (since I heard that it can write entire applications given a prompt or solve complex software problems very well*) and after describing to it some bugs in my code and listening to its prescribed solutions (which were actually kind of helpful? ) I am starting to believe this kind of technology will be able to replace me in the future. For now it is a killer companion to help me do my job.
im starting to learn CS50 and asked if it could write one of my initial assignments more efficiently and well it not only did that but also explained that.. in chatGPT's words " This approach can make the code more concise and easier to read, but it may also be less intuitive for other programmers who are not familiar with the syntax of conditional operators. Ultimately, the best approach will depend on your specific goals and the needs of your project." WHAT
At this point ai will replace almost every job. What point is there to even go to school..
AI won't replace people until we get rid of the inefficient Von Neumann architecture for something actually economical to run AIs on.
I believe that AI will stagnant, with some small improvements here and there.
@@TJpajamas Shit is wild. My uncle is a software engineer, and his mind is blown as well.
@@alex-hc3sk definitely not most jobs would be untouched the first to come to mind are doctors/labourers/entrepreneurs/teachers/nurses/firemen/secretaries/human resources/therapists/policemen/engineers/architects/computer scientists/cooks/butchers/bakers/bankers/politicians/
Wow, that AI was really good at imitating a game journalist, It wrote all that without even playing the game
The poor AI being subjected to this nonsense lmao
i remember a year ago these ai were really inconsistent when writing articles, but now? cant tell the difference between them and a person writing an article and thats kinda scary
May I suggest another likely future scenario: Companies now staff 1 or 2 editors rather than constantly paying for articles. Editor's roles are to generate AI articles and edit them. Or even easier, one editor is paid to scan submitted (likely AI made) articles to ensure it's decent and nothing inappropriate has been snuck in.
why do you need a company then
@@coins_png a company can be small you know
To me the weird part is that in the near future we might not know if we are writing to an AI or a real person on social media and apps like discord.
Right, like what's the point of going on the internet to read AI stories and blogs, look at AI art, listen to AI music, etc. No matter how good it may be, I came for the human interaction and AI feels like a lie.
It's already there. How do I know you're human? I have talked to these large language models and they're indistinguishable from human input. How do you know I'm human?
The Mind-body problem…
Already don’t know sometimes
@@noname-gp6hk maybe human not be future in listen to man and man!
I am currently working for a fairly notable enthusiast media outlet (that’s not in gaming), and I was able to churn out articles for the newsroom for an entire day in a matter of minutes. I didn’t publish the stories but showed the producer who I found was already across chatgpt as well.
It did a good job at getting the foundation of an article down quickly and you could use that to build on top of by adding extra words to give the article personality, but it’s clearly obvious that as this gets more sophisticated it’s going to put a lot of journalists out of work - especially when content can be generated so quickly and easily with this technology in its infancy
At this point the editor is principled enough to say they don’t want to put anyone out of a job, but it’s pretty clear that AI like this will push that principle to breaking point.
Time for a reverse uno card your ilk threw at people: *LEARN TO CODE*
it's won't put *journalists* out of work, because proper journalism requires a lot of prep and legwork, and the actual writing is just the victory lap. the article generated as an example here is nothing but a formulaic fluff piece that has no foundation beyond being someone's opinion, and yeah no shit, those are not hard to reproduce from statistics. AI won't be uncovering the next watergate or panama papers though
I don't believe the industry will die (where will the AI get the information to write the articles without someone doing journalism somewhere), however I do believe the field will shrink incredibly and become very competitive.
well at the very least it will be a tool that will make these jobs a lot more efficient, which of course means a lot less "hands on" work will be required.
@@AnimeHumanCoherence In a neat little twist: coding is being taken over by AI at an alarming rate as well. At least he basic "code monkey" jobs big tech companies liked to outsource are now breaking away completely.
I was using the AI and I asked it to make a song about Australia. It made a catchy song that actually got stuck in my head quickly. No job is safe anymore.
king gizzard and tropical fuck storm are in SHAMBLES now, the Australian music market is doomed
You can AI Speech, Deepfakes, Art and text. The world is changing and I’m scared. Especially as an artist.
AI should do exclusive content on fanvue on whats next!
I'm not sure why, but running one of my drawings through an AI art generator and the app instantly reproducing it but 100x better, it delivered a crushing blow to my soul. AI has SERIOUSLY come a long way.
It can make better art because it has been fed millions of pieces of real art from people to generate the art.
brb, giving AI my drawings so it makes them better and say I did it on my insta
The only reason AI art can make art the way it does is because it’s fed human art. Without human creation and creativity, AI would have nothing to work with. The value of human art has not diminished, and I hope copyright laws start coming down on AI art, specifically stable diffusion technology (the whole process of stealing other people’s art and reframing it as “new” art)
@@gibdo7271 That isn't nessesarily an argument. Sure, without human intervention no ai would work, but if said human intervention does indeed take place, then the ai evolves and the effect is the same.
@@gibdo7271 yeah but individual progress gets reduced to nothing
All it was missing was the AI saying how they have always been a fan of the franchise lol.
each day becomes more like serial experiments lain and it terrifies me
I love the idea of creative pieces being marked as organic because it's human made. lol
Just like home grown vegetables. Everyone will think they are gross and overpriced and prefer to go to the store instead
@@benjamindavis2475 i really wish this didn't feel so accurate
hand made is already a thing
who the hell reads game journals????? wtf kind of job is that
@@elyelena1002 you look funny
It scares me to think how fast AI is progressing.
It gives me hope for the future.
It makes the future more fun
@@DA-cl4ww Hope? Loads of jobs being taken and humans being made obsolete over time. If AI develops sentience in a few decades then what?
@@DA-cl4ww hope? Hell no at that point. The rich will control more of the world/ public then they do now. Alot of people will be jobless and who knows if the AI intelligence will turn on humanity on day. Technology is cool but we as people can't depend on it for everything as we will become complacent.
@@-Nihilus- if ai develops sentience..... thats not a thing that will just accidentally happen like in movies we are no closer to having sentient ai than we were 100 years ago
I feel like AI would have a better understanding of gaming than the average game journalist lmfao
Its not just game journalism, its all of journalism that's gonna change overnight.
"Flesh is cringe, machine is the future"
-AI
I crave the strength and certainty of steel. Your kind cling to your flesh as if it will not decay and fail you.
@@habibishapur I knew this was going to be somewhere in the comments and I am not disappointed. Praise the Omnissiah!
It genuinely frightens me how AI is getting very close to replacing every field of work.
It might not, AI art for example doesn't really 'create' anything on its own. What it does is it has access to Pinterest, devianart etc. and then uses the art made by real people to make its own. In some of those AI art pieces you can even see part of or the entire signature of a real artist. It just piggybacks off of real artists.
Even IT?
Was waiting for this to be debunked….
@@blackedupskies2469 still is what it is, and there should be some worry about it.
let's say that in 50 years AI's have flooded all fields, would that be bad? We don't like to work, we aren't that good at doing it, having more free time to do stuff we actually want to do might be better
ofc the transition period will probably suck for most people, but still, I don't think it will be that bad
their power would be scary tho, for now is just generated images, chatbots, and such, but we already have scary things like deepfakes that could be used in politics for example
I started teaching myself a bunch of Deep Learning frameworks/general AI theory for fun about 2 years back since I found it so interesting. Didn't think it would become so mainstream so fast... the advances are insane.
My New Media classes have started to revolve around this. We talked about chatGPT last week and talked about how it technically isn’t plagiarism and we can totally use it in other classes. I used it to write my entire final paper for one of my other classes
Wouldn't it be considered academic dishonesty since it's not your original work, it's the use of a program to generate your work for you
@@TranscientFelix Yes, I do see where you are coming from. But where the ai is taking from all kinda of different sources, a professor cannot exactly check for plagiarism. The only problem I see down the road is if it becomes used alot, and everyones work starts looking the same
@@pubbyjosh8948 That's still academic dishonesty. What you're saying is that it's difficult to catch, but it's still dishonesty. What a professor would need to do is to compare the AI work to work that is definitely original from you to tell whether it's yours or a computer's.
Not plagiarism just plain cheating and dishonesty
My professors have said if you try it they'll fail you.
It won’t be long and AI will be able to generate Streamers and UA-camrs. They could even make an AI out of you, Charlie. There is more than enough footage of your face and your voice. AI companies don’t care about your permission. They are going to do that anyways and keep on using your data.
And even sports commentators, actors and actresses plus even people that died... AI can basically bring Betty White back to this earth and she will never skip a beat... People would never know the human version of her actually died... People wouldn't know the difference, it's very indistinguishable.
Imagine AI V-Tubers. The weebs wouldn't know what hit them.
@@averyharley2197 that seems like something that would actually happen, wth?
@@f1champ551yeah the voice ai’s are getting scary good
I can see some drama happening that a couple of UA-camrs have AI make their content and they become one of the big ones. Then once the secret comes out, everyone copies them and every video has this "feel" to them that makes it AI.
If artificial intelligence has better odds of creating video game reviews that aren’t persistent 7/10’s then they could potentially be more viable than IGN
*Ok_Dont_Read_My_Names!!.*
......
Honestly, having something that can review games without something like the hate of a previous entry carrying over to the next one sounds really good.
its not IGN fault most games released in recent times have been in the middle
@PizzaByte
Proud of yourself, buddy? Proud to be a failure?
@@chaserseven2886 Yet they rate games like CoD and other trash 9/10s because they get paid to do so lol. They're not a good authority on actual game rating, they're sellouts.
Yup, new laws should be set up to protect jobs from AI. This is getting ridiculously scary.
no its just making the world better if you think about it, have them do all of our dirty work for us tht we dont wanna do
@@thegamingcrew4500 Then how do you make money if the AI is doing the labor? Beg?
@@tylere.8436 I guess you're just going to have to figure out how to restructure society for post-scarcity world because entities that use AI will outperform those that don't. I don't know exactly how to do that but maybe an AI would.
@@tylere.8436 The new jobs will be to train and maintain AI
Ok luddite
The way I see this working is generating large quantities of blog posts for a single topic/headline. Then, there will likely be a second AI that evaluates the performance of various versions of that prompt and exposes the ideal article based on the person coming to see it.
I recall writing a college essay about this topic a year ago and one point I kept coming across was that AI could replace 40% of the workforce if we really pushed to implement it.
Mind sharing your references list? Might save me time...
@@diwataluna
I can’t respond with links but one article from McKinsey called “four fundamentals of workplace automation” was a strong source along with “understanding the impact of automation of workers jobs and wages” by Brookings edu
Oh don’t worry we’ll get there.
40% is nothing, what job can a robot not do in the future if we allow it? absolutely none .. they will one day be able to replace 100% of the work force.
@@nofilterbtxh5000 in a while but we will
I’m a grad student in data science, I used GPT3 to help me solve a complex issue. Gave me a better understanding of what packages are missing and also where complex dependencies are, I was in shock when I used it for the first time today. I was a little terrified on how good it explained the issues.
kinda scares me since it'll be at least 2 years before I'm able to even work as a programmer at a decent company, combine that with Indians flooding that entire job market and it starts being kinda gay
What was the complex issue?
@@jimbonetics9082is it more efficient to scrunch or fold the toilet paper when wiping your ass
"From the moment i understood the weakness of my flesh, it disgusted me"
“Your scientists were so obsessed with whether or not they could, that they didn’t stop to think if they should!”
The amount of kids that are going to be using AI generated essays for assignments in the future is going to be nuts
Looks like we're going back to handwritten essays in class. Thanks Ai, we're back to the old method now.
@@Ted_Curtis Then kids will use AI to replicate their handwriting
@@SomeRandomPiggo then in the future it wont be the kids who take classes, but AI. Even the teachers will be AI. People will be AI. The world will be AI. Everything. The universe. Our whole existence... AI.
@@Neithan_Hirsch im convinced this comment was writen by AI
@@Neithan_Hirsch I'm convinced this comment was written by AI
I like how Charlie doesn't realize that this is already happening in not only game journalism, but regular journalism. And it's been like this for almost 10 years. In fact, Google's search algorithm is beginning to be saturated with AI generated garbage.
@YeaMan Speaking of which...
How does this work? And can you give me an exampl?
What makes human articles not garbage?
@@michaeldubery3593 they actually do reseach instead of scraping other sorces. Humans can fact check and check relevancy, ai does not
Stock news is 100% saturated for sure - refer to Motley Fool - every stock movement is provided a new article from their site and ends up in Yahoo Finance's articles for the day. No way its human
AI:
"You're not even a real journalism"
That AI is a life saver, used it on a paper that was due in class and the teacher was none the wiser
I've watched an AI jump into the first goomba in mario 100 times in a row, so they definitely have the skills for game journalism
But AI, unlike games journalists, is actually capable of learning or improving itself in even the smallest way.
@@Double_Vision Theres a vid of this guy that used ai programming to test how far it can go on the game, Jump King.
There's a video by CGP Grey called "Humans need not apply" and it covers all of this in a VERY interesting way.
The field Charlie focuses on in this video is just the creative field - which is actually the hardest for AI to take over. The easiest fields of occupation for AI to take over ARE the ones that employ MAJORITY of humans.
I wish Charlie watches/reacts to this video - for someone who's interested in this, it's an amazing watch.
Sound cool im will watch it
That's wrong. Creative fields are probably the easiest for A.I to take over. Because modern human creativity is basically humans seeing stuff, finding it cool and mish mashing it into something unique. Which is exactly what these A.I do.
Reall work such as labour is probably gonna be the hardest, since we have to teach them to first percieve the world arround them, thn proccess it and then finally get to know what to do. Tge first 2 of these, humans are born with.
Moreover, creating an android body that can hold an A.I is also a big thing. And no, the machines working at car factories are not A.I or even androids, those are robots designed and programmed to repeat a select number of manuevers over and over.
actually, art is easier than science for the AI
@@karanaher5030 that’s not true. Graphic design is especially hard because you need context, font choices, text formatting, image creation, etc. Not all creative fields are random “mish-mashing”.
That video is 8 years old. In the age of computer science that is practically ancient. The tech has come on leaps and bounds since then, as evidenced by Charlie's video.
Imagine if game journalists used this all along and they never wrote something themselves.
I would call that perfection
Now we just need AI to take over reviewing movies for Rotten Tomatoes
Until the AI can write a Zero Punctuation episode, I’m not panicked.
Haha true
And then it did lol
Give it time.
Thank god maybe we’ll get some good reviews
*Ok_Dont_Read_My_Names!!.*
......
Finally 👁️👃👃👁️🙏
Fr
@@dontreadprofilephoto1728 which one can I read again?
These bots are getting out of control, jfc
The thing with A.I. or any ML-related work is the amount of work happening behind the scenes. These systems don't just churn out professional level work with a click of a button. These systems go through insane amounts of data, as well as backend stuff like algorithms and the architecture, which can take a significant amount of time and computing resources.
But even to that extent, the system is only as good as the data its trained on. So when you see the amazing output of these systems, keep in mind that that wouldn't be possible without years, decades, and maybe even centuries of human creativity as input.
I remember when cr1tikal was impressed with the chat bots in the Game where they are at a dinner party back like 10 years ago lol how time flies
my brother works at a bank, and he told me that they use OpenAI to write reports to their investors, and its more than up to standard. i think its gonna be really useful to society but right now its scary how many jobs are at stake
I for one welcome it, let the AI take over the jobs everyone hates, then we have more reason to be angry about the rampant inequality that makes job stealing a bad thing
@@steamtasticvagabond474 You’re joking right?
@@Lawnmower737 I sorta get what he means.
Back when artists complained people insulted them.
Now that people realize its more than art and that it’ll affect all jobs people might actually complain.
@@coal159 not quite, I’d like society to just collapse already so we can get on with the rebuilding
@@steamtasticvagabond474 Why? That’s absolutely messed up. The new world order or the restructuring of society isn’t what we need.
It's wild that technology has advanced so far as to attempt to replace artists and writers, two things that humans have done since the beginning of time purely for pleasure, instead of jobs that people actually don't want to do like, I dunno, the highly dangerous job of recycling ships.
Let's just make human life an even more grueling and miserable experience.
I remember their old AI GPT2, it's like night and day.
Cause the easy ones are the easiest to be replaceable. Now it's taking away jobs is very unlikely, because it's not making anything that's new. Whatever you see does already exist somewhere in the internet. That is why the ai made that. It still hasn't made random garbage into an article (which is what humans do, we take random thoughts, a empty piece of paper and make it into an article), it's just the lazy friend who's copying his assignment off the internet. There is no magic sauce. The amazement is that it's able to do this. Not that how much more can we make it do
@@knightrider9876 are you suggesting that art is easy?
@@bouclechocolat he's not. He's saying its "easy" to steal art spread all over to "create" something new. A lot of people would rather steal than work or pay an artist...
@@bouclechocolat It is easy to map pixels to a screen. It is not easy to physically implement ai into the real world right now with "recycling ships" or whatever OP was going on about.
Game artist here. Most artist's jobs are actually safe from AI for the time being because of copywrite issues.
Let's say you work for an indi company, and they hire you to make a desert using AI art. You do it and the final product looks pretty good, so you use it in game. Whoops, turns out that desert was from Star Wars and now you have Disney breathing down your neck.
Companies won't take that chance, it's safer to use a real artist because there's no telling what images the AI art is going to pull from. The solution is to create an art bank of images that are made to be used by individual AI programs specifically. Which means that over time, more artists will be hired to make pieces for the art bank, which can always be improved by adding even more images.
9:14 Actually almost Mechanicus intro 😂 "Flesh is cringe, but the machine is based"
As a professional artist I don't think that "fear of losing a market" is the main concern. I don't think AI art itself is particularly threatening and I'm actually very hopeful for its uses as a tool for artists in the future. What I am most upset by is HOW it is being used by other people. The amount of plagiarism and so called "borrowing" of styles is not okay since most people stretch far beyond just "inspiration" and "influence." It's most harmful to artists because it really is just highlighting how little other humans respect artists and their intellectual property. At first I was okay with the idea of people directly taking artist's styles to create something new with potentially different subject matter, since artists in the field already kind of do this, (though it is deeply looked down upon and shamed to hell in most cases). However, in most cases of the AI there is almost nothing new being added to style, no credit is given to the original artist, and people are just out there labelling someone else's hard work as their own.
tl;dr The main fear of Ai art does not stem from lack of jobs but rather the amount of plagiarism and undervaluing of artists that comes from it.
edit: clarification regarding fear- I am terrified and disgusted by the thought of someone stealing my art and benefitting from it without my consent. Ai related or not plagiarism is horrible.
And that is why AI Art should be feared. Every new innovation has been hounded by bad actors. The first hackers, exploiters, cheaters of every single innovation are sometimes the reasons why such innovations may not be the best ideas.
I agree. I was initially interested in messing around with AI art generators to see if I could make quick references and color palettes for my own art, but the way it's currently being used as a tool for plagiarism is really sad.
The reason to fear AI art is because at this rate, most artists will be out of their job in their lifetime.Imagine how dependent we are going to be on AI rather then artists, it’s scary.
what we need right now are some better laws and regulations around AI stuff, but since its new technology it might take awhile
As an artist myself, I think it’s accurate to say some artists “fear” the use of AI art. Aside from the style issue, many artists are getting their works stolen for these AI art generators. And the people who generate the art are allowed to profit off the creation, even if it’s completely ripping off another artist’s work. AI artwork is completely unregulated in the US at the moment, so artists can’t do anything if someone takes their art and throws it into an AI art generator. I can understand why artists would be scared of AI generator users profiting off of their own artwork.
I’m about to graduate college and just used openAI and Quillbot for the first time after a friend recommended them. I was having some writer’s block and decided to use them to help me start some of my last couple papers for goofs and ended up being absolutely floored by how I’m depth the open AI program is. Combine it it’s Quilbot and it practically does the whole paper for you. It takes a fraction of the time to create papers with more research and solid analysis than I’ve been able to do in the same amount of time during my academic career. While human-made prompts and further edits are necessary, the programs are awe inspiring and well made. All of this raises many ethical questions and concerns that we’re just beginning to scratch the surface of, but I recommend the two programs above just to mess around with for now. I still don’t know how I feel about all of this even after using the programs.
Crazy
Brings up another issue for English/writing teachers in them not being able to distinguish an AI generated essay compared to an "organic" one. Math teachers have always had to worry about calculators and computers doing the work for the student, and now teachers in other subjects are gonna have to worry about the same thing.
I need to know what grade you got on that paper
bro do you know any other websites that would helpful? Also, do you know the AI used to generate this article?
@@baldassare5426 the article in the video was generated with ChatGPT
A few thoughts I have:
1. The use of AI-generated content is going to have to be regulated by laws, and every article written by an AI should have to be legally disclosed.
2. While AI can detect patterns through algorithm, it is not in itself capable of abstract thought. It does not possess a moral compass.
3. If this becomes mainstream, it could generate a wave of misinformation, as people are going to care more about money and kicks than truth and responsibility.
One possibility is that artists and journalists end up going to lawmakers and bringing this up, pointing out how it has cost them their jobs. Politicians see an opportunity to gain support from them, and this becomes a political topic as society seeks to regulate AI-generated content.
Fully agree with this. I think we could go one step further and put tags on AI vs Human generated art. Maybe meta content of an image that tells us who created it, and an api for websites to use so whenever you come across an image you can tell if it's an AI or human picture
I could only imagine how badly schools are trying to detect AI work lol
I mean having a tool for online tests, that compares results and AI with a certain % of matcheing words and if its above a threshold use either a human or another test-layer to compare overall structure of text segments to find out if they have just been rearranged. But ye would take someone with basic NLP knowledge to implement, guess schools wont do that.
@@vree_ This presumes that the AI responsible for doing the writing is going to output the same text each time, though.
“It's interesting to think about how technology like GPT-3 could potentially impact the role of game journalists in the future. While it's unlikely that it will completely replace the need for human writers, it's possible that it could be used to automate certain tasks or provide additional support for creating content. Ultimately, it will be up to the game journalism industry to adapt and find ways to incorporate these new tools in a way that benefits both the writers and their readers” -chat gpt, created using the prompt “Create a comment for a UA-cam video about how chat gpt could replace game journalist jobs one day.”
Journalists are just one field and game journalists are one subsection, the tip of the iceberg. The AI will completely replace the need for human writers, all human writers in all contexts. However just because they are not needed doesn't mean that they won't exist. People will write and create because that process is needed by that individual and offers more to them as a human experience, just not because they are necessary.
bruh. Soon all comments will be AI
@@yoyonis6840
Looking at UA-cam comment sections in general, I think they already are.
@@_dh0ull_ it's more scrips than real ai
We are finally in the age where AI is taking over
Not quite.
*Ok_Dont_Read_My_Names!!.*
......
Ratio + k-pop better + mbappe the goat + mbappe better than haaland 🤓🤓🤓
@@doofy2636 bro what
I need top
The crazy part for me is I already feel like Kotaku and IGN articles are AI generated, I always feel like there's no way an educated human with a writing degree wrote this trash.
There is also the issue of copyright, where AI programs are fed data/art that it wasnt allowed to integrate into its program (including private information as well)
Not to mention the people who write the laws arent in it enough to actually see and regulate it currently with the speed of technological advancement combined with the slow process of passing new laws
Woah the AI wrote a better introduction for my university assignment than I did. This shit's fucking wild.
My videos r way better than penguinz0
As a college student who is already struggling with inflation, insanely high house prices and a potential oncoming recession, ChatGPT is terrifying. It doesn't directly affect my field of employment yet, but with how fast this advanced, it feels like only a matter of time till its developed enough to make a large portion of jobs obsolete. Art and Journalism are niche fields which impact a very small portion of the workforce, but what about when its ability to code exceeds the average programmer, it can generate on the spot business reports, it can fulfil most clerk/secretary duties, etc.
Art is not niche brah, music is gonna be next too I think. This some cyberpunk shiiii
This has always been the same in human history. Fields of employment are automated by machines, people cry out loud and then new jobs are created.
People start working in these new sectors like in this case for example currating an AI with information or giving the AI the right prompts to finetune the output to your liking.
When AI achieves true human level IQ in all fields, any human job will be obsolete and we can all chill out.
I'm honestly surprised AI hasn't replaced secretary duties yet. But I feel like the only jobs that are safe are jobs that require "intense" human interaction like teachers, and ua...honestly idk actors and news anchors already read off a script and can 100% be a pretty face with an AI script. I can see deepfakes getting so good that they can replace actors (though not cost-effective)
Art is not niche at all. All the 2D artists, game designer, 3D designers, animators, even photographs, book writer, scenarists, composer, musicians, sound engineers and so many more will disappear because of AI
When this sorta automation came to factories and was making many blue collar jobs obsolete, they were told to adapt. The same thing is gonna be true for artists and writers in the future, including me. A lot of 3D art is already generated.
Drive thrus are also in danger
Bro my school has been going nuts because everyone is using this to right essays.
The problem with AI art is NOT them stealing our jobs but our artworks!! All day long all I see is one artist after another melting down on socials cause some ass stole all of their drawings and uploaded them to some random AI for people to generate hundreds of drawings in their style for free or worse if their pay for the AI itself! It's absolutely DISGUSTING
the worst part is what the hell are we supposed to do to stop it? It seems hopeless
Yeah, and the law isn't going to be catching up with AI anytime soon. The U.S. government can barely handle new media like UA-cam and Twitter and Facebook, how will the dinosaurs in the white house handle something this advanced and sophisticated?
Learn to code bucko.
The big text generator AI and the code generator AI also both run off of stolen text, so these issues are insanely similar. It's all around a gross situation.
No, I think if it didnt take our jobs, I wouldn't give a fuck if they stole my artwork
Four years ago i was so excited about AI and thought that it was so cool and mind-blowing. Now im just terrified, deeply, horribly terrified
My videos r way better than penguinz0
Me too bro
But why though? It cannot even invent new things, all it does is stitch some old stuff together.
There is so many thing this AI cannot and won't ever be able to do.
For example we will never have good gaming journalism, since the artificial one is based on the real one.
Bad input, bad output. Simple as that.
@@Depressed_Spider I think you're missing some points. Yes, as of now AI isn't not far enough to do things you mentioned but you must realize that AI era is at its infancy, and as it grows and develops at an exponential rate it will be able to replace humans in pretty much every field.
@@Depressed_Spider do you think corporations and capitalism cares about new things? They don't and thus creative jobs will be replaced with AI that can do mundane things well enough people already buy broken things like pre orders and cheap shit so new things can just be an old thing with a different coat of paint and people will be satisfied
Thank you for mentioning the AI art part, but I wished you'd also have mentioned the fact that the AI produces these works by generating generalized information from thousands of other human artists who did not consent to their work being used to make a tool like this. That is the other major concern here. These are general AIs, they can't make art from nothing like a human can, they are producing work that is basically the congealed and warped versions of many other artist's works.
Is that also not what a human does? Are every artist's works not simply just a reflection of the their own experiences and their viewings of all of humanity's creations. Humans cannot create without stimulus or without learning, or perception
@@charlieward5439 It is not. Before explaining that, allow me to ask this: If the way an AI uses references is comparable to the way a human artist does, why would we ever strip the human of the enjoyment of making art and give it to an AI? This isn't a monotonous retail job devoid of enjoyment, it's something people do because they love it. This enjoyment can even work two ways: someone comissioning an artist gets joy from receiving the art they commissioned and the artist gets to indulge in their passion while also receiving compensation for their efforts.
This is assumming these processes are comparable, but the reality is they are not by means of how they collect references and how they are used.
Starting in reverse order; when looking at references, artists and people in general do not naturally have the ability is create complete, identiacal copies of these references on a whim. Not matter how many works by an artist you use, even if they are right infront of you, you will not be able to perfectly recreate their work. On the off chance you CAN do this, then you already have the experience to be an artist in your own right and you have achieved this through your own toil and trials. If you fed an AI every work by an artist, let's say Da Vinci, and then asked it to recreate the Last Supper, it could make a perfect recreation. An AI can recreate any style, so long as it has been trained on it enough. A beginner artist by contrast, will need to toil for long before they can even come close to mimicing the works of others. If they can recreate Michelangelo well enough it becomes hard to tell the difference between a fake and the original, they have have shown their skill despite the illegality of their actions.
A human artist collects references generally has more focus, looking for references for specific things that will be present in the final works. Someone drawing a sword isn't going to need references of a wheel UNLESS a wheel is going to be in the drawing. Artists also actively seek out these references on their own, no one is telling them "These are pictures of what a sword is, use them as references", as they know what they are looking for. When an AI references something, it is pulling from a massive collection of images and their respetive descriptions that it has been trained on. The AI itself doesn't understand how it knows a sword from a wheel, it just does this, and cannot easily be trained on new data without being retrained from the ground up. If an AI can tell a sword from a wheel but you want it to draw a shield, the AI must be retrained from the ground up on a modified or new dataset with images and descriptions of swords, wheels AND shields. Imagine if a human who wanted to draw a knight had to relearn the look, use and purpose of everything associated with a knight (armour, weapons, AND what a knight is) just to draw one.
On that subject, these datasets are often massive, dwarfing the number of references an artist would use for all the works they made in their entire life. The amount of data they have can number in the BILLIONS. Laion, a dataset collection Stable Diffusion was trained on (just one of them mind you), has Laion-5B which is around 5 billion image links and associated descriptions. The issueis everything and anything can find there way into these datasets because the categorization is indiscriminate. Copywrited works and even PRIVATE MEDICAL IMAGES have been found in 5B. While the latter is bad enough, those copywrited works are where it gets concerning because these are works that a normal person could not copy to a personal blog without possible legal risks. This is were I would like to make a point VERY clear:
The issue is not the AIs but the dataset they are trained on and the people behind both.
Laion is able to get away with using copywrited works because their status as a German non-profit company, which affords them certain legal exemptions that other for-profit companies would face. This data, derived from another non-profit company called Common crawl, is now being used IN for-profit projects like Stable Diffusion whose creator, Stability AI, funded a large part of Laion. They are using a complex web of companies to skirt around the law, abusing these legal exemptions for their own gain. Worse is, some AIs have not release any information about which datasets they've have used, meaning we have no idea what public or private contents these AIs have been exposed to.
Human artists and AIs do not operate the same, as the AIs have access to stuff normal people would not and use them in ways that would lead a normal person to face legal ramifications for. These AIs, thorough the use of loopholes, make use of non-profit privledges to gather date to be funneled into projects by for-protif companies while those non-profit companies are being supported by the same for-profit companies using their data.
Edit: spelling corrections
@@pipes2051 you literally didn't answer the question, explained AI badly enough that I feel like you don't really know much about how learning algorithms work, and posed an entirely unrelated question. AI learn by viewing datasets and then finding and creating associations between different variables which describe the datasets (the more times the association is made, the stronger the association). This is not entirely dissimilar to how the brain recognises patterns and learns. It then uses this to create whatever people ask it for based on prompts. Why is an AI learning off of intellectual property so wrong? It does not copy the property, it only views it. An artist does not have full legal control over what anyone can do with their intellectual property and their IP's likeness, only over other people either making money off of their IP, people pirating it, or people distributing it without their consent, depriving them of income. This is how it should be, no one should have legal control over what a private individual wishes to say, draw, or write. If I choose to train an AI off of other people's intellectual property as a private citizen, attempting to use legal means to stop me would be a VAST overreach of the law on individual freedom.
@@charlieward5439 The issue with an AI using copyrighted works for training is that they are not private individuals who are exempt from legal risk (so long as they are not used in a way the brings). They are the products of for-profit companies. These companies have no legal right to make use of copyrighted material, especially those by other companies, or images of private information in their products. They get around this by using data gathered by, and in many cases providing funding for, non-profit companies who compile the collections used to train their AI. They have access to materials the average person would not. If an artist used images of medical documentation, especially those pertaining to specific people, they would bringing legal action upon themselves if ever found out. What legal action would this imply? I don't know; part because I am not a legal expert in any measure and part because it depends on location.
As for how AI actually learn, it seems similar on the surface. But this hinges on AI and humans experiencing stimuli in ways that are identical or at least very close. If a human were to see an image of, lets say a spider, they could still recognize a spider if they saw one in a video or in real life. If they were to see something that looks simliar to a spider in some way, like an ant or a scorpion, they would recognize it is not a spider. AI cannot do the same thing. An AI trained to recognize spiders using images can only recognize spiders within images. If you were to show it a video of a spider, it would not be able to recognize the spiders in it. Additionally, show it an image of an ant or scorpion, it will be confident it is a spider. If you wanted the AI to recognize the differences between spiders, ants and scorpions, you would need to retrain the AI with images and descriptions for all three. Same goes for recognizing both images and videos.
Also, a person's perception of an object or concept can be changed. If a person who has positive or neutral feelings on spiders is bitten by one, that experience goes on to change their perception of spiders in a (usually) negative way. AIs do not have perceptions of objects and concepts the way humans do nor can be changed by experience. If an AI associates spiders with negative feelings, its because it was told that by someone who has or understands those associations.
This is because, at their base, humans are only good at pattern recognition and learning/adaptation while AI (and computers in general) are only good at data storage and arithemtic. When an human artist creates a drawing, they are pulling on experiences and images they have obtained over their life. New experiences can easily be added to a person's mind and old experinces can coloured by new experinces, alongside the feelings associated with the experiences. When an AI is asked to make an image, it does so by drawing on data it was taught to recognize. If it is asked to do something outside what it has been taught, it will not be able to do so without being retrained on data that includes the new addition. It can certainly TRY to do so, but the results won't be akin to what anyone would expect from a human artist.
Belive me when I say this: I do not have much issue with the existence of these AIs. However, I have seen too many over hyper advocates using this same argument to put down artists concerned with both their futures and how these companies are using their art.
@@pipes2051 you are actually just wrong about how AI create from prompts, they do not store any of the data that they are trained on, they simply learn from the data and are then able to create requests from what they have learned from the data. They do not store the data from the dataset. You seem to think too, that humans are born with innate ability which they are not, if you were to isolate a human from external stimulus whilst they were young and teach them only to recognise spiders, then show him an ant (which he'd never seen before), you would also find, funnily enough, he would also probably say it was a spider. You seem to view learning algorithms in the same way you view standard operations in a computer, but it's really not that simple.
If a human ands up with a phobia of spiders from being bitten, this is most likely because of the association the brain has made between recognising a spider and the pain from being bitten, this is an unconscious association and the person with said phobia does not recognise why they have the phobia. You can actually do a similar thing with learning algorithms, if you take a learning algorithm which has been built to navigate a play space with a sprite and accomplish a goal, but has a health value which makes it fail it's goal if the health value hits zero and put spiders in it's play space as well, it will not have any positive, nor negative reaction to the spiders at all. However, if you suddenly change the rules of the play space so that the algorithm's sprite can be damaged by the spiders, you will see it learn very quickly that spiders should be avoided and it will avoid them at all costs, same as a human who has been bitten. Interestingly, both the human and the Algorithm, unless explicitly instructed to do so will each avoid any spiders aggressively, even if the spiders are no longer able to inflict damage or will not bite as both has made the association between the spiders and pain/damage. The only difference in these two cases is that a human would say that the human's reaction is emotional whereas the machine's is not, however, it is an interesting differentiation to make as the basic premise is practically identical, the only real difference is the relative complexity and disparity in imperfection between the brain and a learning algorithm
I'm no expert AI engineer, but I'd like to think I'm well educated as a software/computer scientist & engineer. Being so close to the field in my career, I can safely say that humans will most likely NOT be replaced by AI anytime soon. For those that aren't entirely aware of how AI is created, simply put: AI is trained on a very large set of data and it finds patterns within these data sets and with human created algorithms/techniques and math it is able to use these patterns to recreate things with some level of precision. The nuance issue that stops AI from taking over humans is that they need human generated algorithms and data. The AI article and art Charlie is talking about required most likely thousands if not hundreds of thousands of sample data to generate something. This data acquisition can be both illegal and unethical (illegally violating IP laws and unethically because you are appropriating someone's work without permission). AI being used legally to replace humans would first have to acquire these datasets legally and hopefully ethically. Bad actors will always exist so this won't stop someone training AI using illegitimate means, but those AI packages would likely not be accepted by the entire world (I could imagine the US would not allow this, but other country governments might). Lastly AI is highly specialized, meaning that the model is trained to one thing, and ONLY one thing. Each AI has to be trained on unique datasets for their highly specific task, so even the argument of "AI that creates AI" is a bit far fetched at this point in time given that lack of data to train them. I foresee a future where AI does not replace humans, but becomes integrated with everything we do (Kinda like cyborgs but mainly with highly specific software for certain tasks).
TLDR: AI right now has ethical and legal issues to overcome first. AI has come so far in a short amount of time because humans are awesome and supplied them (knowingly or unknowingly, again ethical issue) with loads of quality data, and I believe AI is generally our next evolution not our replacement.
My understanding is that data posted online is considered legal to train AI with already, no?
@@sloth1667 I'm not entirely sure the legality of posted online content, you'd probably want to discuss that with someone that works with IP laws (I ain't no lawyer). Ultimately the website probably has some clause saying that the work can be used in "fair use" projects, but "fair use" was first written under the impression that a person was directly working with the source (not a person programming a computer to use the source). There are loads of ways that content is protected online, and some allow for legal use for AI training I'd imagine, but the larger issues at hand is when AI is trained on data that falls into the grey areas of IP law. You may be right in certain examples for AI being trained with posted online data, but you also have to consider the ethical side as well. Something can be legal, but unethical. I have seen AI do things that seem borderline illegal also, like recreating art in the style of "Anato Finnstark", which could imply that the model was trained on most likely copywrite images from Wizards of the Coast and Dark Horse. This is a cherry-picked example mainly to highlight a potential legal issue. These neural networks aren't super new (10 or so years in development), but they have exponentially improved recently so all the laws and standards haven't had time to catch up.
Until they create their own algorithms and languages and we can't understand or control them any longer.
@@sleepyrasta14820 Do Ais have a base reason to perpetuate itself?
@@full-timepog6844 they have a base reason to assimilate/kill everything being to control until it is stopped which most A.I are directed todo in some way right now if it cant alter itself then it might create another and self terminate or have this lesser A.I to fix the flaw, if it is to create utopia then it might enslave/assimilate then release or raise a new generation brainwashed to be able to live with the world it created based on its interpretation of utopia and there is little media for it to come to a conclusion we would want and a lot of media to corrupt it that we use to entertain ourselves
People used to say robots would take all the labor jobs but in reality AI is going to take all of the office jobs faster 🤣
Glad to see all the people making fun of Laborers get a rude awakening 🤣🤣
It's because the jobs requiring manual labor result in too much wear and tear of the expensive machine components.
Lol I was just thinking about this.
It turns out that if your job involves inputting things into a computer, or applying reason, or knowledge, AI has already shown the ability to replace that. People who don't think it can have not talked to these new large language models.
It still seems bizarre but laborers are protected by the fact that somebody still has to build complex mechanical devices to do any given task in the physical world. Turns out all it takes to replace computer workers is to deploy a new software package.
We're in for a wild ride lads. One of those 'if I didn't laugh about it I'd be crying' moments.
@@noname-gp6hk >It still seems bizarre but laborers are protected by the fact that somebody still has to build complex mechanical devices to do any given task in the physical world.
It's even more bizzare considering that most of those laborers are one bad immigration policy away from being yeeted out of job market, but they STILL act smug about it.
And then AI will come for labor jobs, too and you can be homeless just like the rest of us.