This Japanese Author is in BIG TROUBLE
Вставка
- Опубліковано 8 лют 2024
- A.I. is growing at a massive rate and is used by millions... but what is the "correct" way of using it?
Edited by Luke: / lukecraigphoto
#theanimeman #joey #japan
SUPPORT ME ON PATREON: / theanimeman
MY CLOTHING BRAND: nonsense.jp/
Twitter: / thean1meman
Instagram: joey.the.an...
6:57 "Kudan explained later in an interview that she only employed Al-generated text in the responses given by Al-built in the story." Alright, so I got clickbated by this video's title, and Joey is clickbaited by the news article's title.
Yeah, but he also used this as an opportunity to open a discussion, so I’m not mad.
So far, the best use of AI I've seen is for memes.
Uh... maybe you should read up on it a bit.
@@matthewalvarez6884 idk about that, the DC joker and batman story was pretty funny, the last one was especially a killer
pron
Try using it. Get some ideas from it. If they are good for memes, wait until they are at least 5 years from now. It's not harrowing. It just is.
@@Zen-zt4uk lol this was written to look like AI. It was interesting to see. But I need to know exactly what is the one you're talking about, because then I can show receipts about it.
6:50 Wait, the article says that she uses AI as a reference to AI-specific responses in her story... because her novel is about AI, I don't see any issue with this. The AI didn't write any part of her story, only referenced AI-like responses 😅
i think her use of AI is totally sensible. it’s like interviewing and studying a hair dresser if you’re writing a story about a hair dresser.
Proving the adage "don't ask for permission, ask for forgiveness."
This is a special case where the usage was conceptually and logically appropriate. This proves nothing in regards to "don't ask for permission, ask for forgiveness." Take a look at AI generated art and try telling me that you should 'ask for forgiveness, not permission' again
Do you know *HOW BADLY* they fked up by not asking for permission and instead tried to ask for forgiveness instead?!?!! AI generated works are basically ineligible for copyright for a number of countries, which also means they can be ripped from wherever for basically little to no consequence. It also means you cannot claim a work is your own when submitting it for contests or publication. And that applies to basically *ALL* AI generated visual art. All of this is because a number of AI art generator developers decided that scraping data/images from the web was 'perfectly fine' for whatever they were developing -- and it was not. They have lawsuits on their hands now. And this severely impacts *ANY* AI art generator going into the future, as even if they *do* decide to go through with the proper and ethical method of getting permission - all of those are now completely screwed over since generated artworks are not eligible for copyright
Should I also mention that legitimate artists would very much use AI art generators if they were actually made ethically? Yeah, a huge amount of artists will basically never get near one of those generators now, not even with a 10 ft pole. This isn't even considering the legal rulings where AI generated art became ineligible for copyright protection. This is an issue of ethics and what type of a statement is made for even using AI generation in the production of their works. The most it is allowed to be used for is inspiration, and even then artists take issue with that since there is lack of proper documentation and reference. "Art doesn't exist in a bubble" as one of my professors said. So where is the connection to other works of art and existing social circumstances? Yeah, AI generation doesn't care about any of that
Yes, tell me again "don't ask for permission, ask for forgiveness." Go on. Please. Go screw over any number of other people and future ventures into anything by deluding yourself that everything is going to be perfectly fine. AI art generator developers should have *KNOWN* that they were violating any number of copyright issues doing what they were doing but decided 'fair use' was going to be enough -- it wasn't a good enough defense and could never be used as a validation. And they clearly understood *nothing* when it came to established rules and procedures in art
And that's how most of the bad things happen and will continue to happen.
As a Literature major I think her idea is pretty interesting and possibly revolutionary (depending on how amazing the story is), but I would probably integrate it better with the marketing and promoting and use it as a selling point which would probably, I'm guessing, allow her to equally win the prize without having this "problem" arise. Using an actual AI's answers for an AI character sounds pretty straightforward to me. In a literary aspect it can open amazing and productive conversations and new studies going forward.
Sorry for going off topic- but I’m about to be a Senior taking AP Lit, what 4 novels would you recommend? I read Memory Police that was good but literally don’t know anything else 😭
@@rae7080 I'm so sorry 🙏 I live in a different country so I don't really know what to recommend for high school. Maybe you could ask your teacher for a list of novels to choose from or ask another student who has taken the class already. UA-cam videos or a quick Google search might help too but better yet try asking an AI, you got nothing to lose
As a visual artist/graphic design student, idk, AI can be a helpful tool but also a curse bc it can steal your art, people seem to use AI than conmission artist bc its "cheaper" and takes less time etc etc. At school students make essays with chatgpt and teachers get worried that kids won't learn irl if they just ask an AI to do the work for them. Honestly idk what to feel about it bc I do see both positive and negative stuff in AI so yeah...
AI is just a tool and in the end I think it comes down to raw productivity vs personal skill. In most cases productivity will come out on top because that's how it is. In school the building of personal skill should be the main focus and the only way to ensure that AI isn't abused is to test those skills in ways that makes cheating impossible.
I really don't like the art stealing argument. AI doesn't steal art any more than a real artist being "inspired" by the style of another. It is basically a system for recognizing what makes something what it is, connecting that concept with a description and then being able to reproduce it on command. It's an "artist" that has looked at millions of pictures while reading descriptions of them. The material it has been trained on is not stored anywhere so it can't really "copy paste" anything, it can just understand the instructions given to it and follow them based on its learned understanding. Simplified it's image recognition with an extra step. Instead of looking at an image and saying "Yes, this is a cat" it creates an image of what it recognizes as a cat.
The AI revolution will increase productivity a lot. It will make all forms of creative work more accessible to a large number of people by lowering the skill requirements, which in turn will create an enormous boom in the creation of all forms of media. Small teams will be able to produce things that previously was only possible for the largest studios with their seemingly unlimited budgets. Those large studios will also adopt it to create products that were previously considered infeasible even for them and are still out of reach for the smaller competition. For the customers this will be great.
The downside is that the value of personal skill within the field will go down a lot and competition for work will become harsher than ever. Art is just the first field of many that AI will conquer. A lot of people will be impacted by this but in the long run, for society as a whole, I think it has the potential to be one of the greatest technological advances ever.
Honestly, the use of AI with students in schools falls on the teachers and what expectations they set. The solution is simple. When we were all in school, we had to provide our sources and footnote anything that came from an outside source. If a student takes 3 minutes to prompt an AI platform for an essay, there'll be none of that...and the student would (or at least should) get penalized appropriately on the mark they receive. However, if it then forces the student to seek out sources to "fake it" -- well, they're still learning and reading and doing the work, albeit at a base level.
Think of it like writing a cheat sheet for a test...the joke is actually on the student that thinks they're cheating by going through all the material and writing out the important aspects for what they think will help them on the test. Well. Funny enough, that's what learning is lol Cheat sheets have their cheeky merits.
Something about AI art is that they're too "perfect" at superficial level. AI can help artist with tedious process, such as drawing thousands of tiny hatch lines, a crowd of people, or complicated perspective, but it is up to the individual artist to edit flaws and unnecessary details to show what he/she really wants to express. So lazy AI art is this "perfect" looking picture that lacks the individual artist's unique idiosyncracies and mental state. AI can emulate style of thousands of artists by gathering data out there, but human artist changes and evolves over his/her life time in unpredictable way.
@@SNixDThe idea that image generation models are inspired in the same way humans are isn't accurate. The companies behind these models want you to think it is because it's good marketing, but it's just not the same. Whether the models are stealing art or not is a bit more complicated, and I don't really know the answer.
I think most programmers know that ChatGPT is very limited in how it can improve productivity in programming, despite all the hype around it supposedly replacing programmers. It's good at coming up with a basic template, one that's probably similar to its training data. It's ok at explaining concepts (but sometimes just makes things up). When it comes to generating actual code, it makes stupid mistakes that a human would never make. I've had situations where I would tell ChatGPT exactly what it did wrong in its code multiple times, and it would change the wrong thing in the code generation. It just doesn't understand what it's doing, and why would it? It's not trained to understand things - that's impossible with today's technology. It's trained to imitate human understanding.
Like any tool, AI shouldn't be used without permission. End of story.
Using ChatGPT to get ideas is almost no different than getting ideas from other people. In my opinion, it's probably worse since ChatGPT is better suited for giving logical answers than creative ideas, so I don't see too much of a problem in that sense. But it obviously isn't ok if it's writing the story for you.
I appreciate her honesty
"Why did she admit she even used AI!?"
Well this is the funny thing; because if you have read the book that won her the award her intention is pretty much to simulate how an actual AI would "think" by asking them directly and writing up what they answered basically word per word
Edit: shouldn't have paused while typing this bcs not 30 seconds later the article literally mentioned it 💀
*cough* 6:55
@@voccapoei read the whole comment before you reply
@@mumtrz Your comment is a mistake and should be deleted. But you like those likes too much. 🥲
@@voccapoei your life is a mistake :Low_Tier_God:
@@mumtrz Coming from the guy who rrally wants to keep those 94 likes. Sad
Ngl I use Google to search words up just to make sure that I spelled it correctly
That is not going to be a good win; given that Alphabet Inc./Google use so-called "stemming" algorithms to guess at spelling errors.
Use a g.f.d. dictionary, please.
@grey5626 okay Mr perfect grammar 👌
Using AI to generate a story should be disqualifying for any sort of award as a matter of course.
She didn’t use it to generate a story, she used it to simulate real AI interactions and responses for her book.
How about an "AI assisted" category, so they're not dropped, but grouped together?
Do we need 'CGI assisted' category for movies?
Do we need 'Photoshop assisted' category for photos?
Do we need 'thermomix assisted' category for dishes?
Do we need 'auto-correct assisted' category for my comment?
Or do we care about the end result, not the tools that were used to achieve it?
@@kukuki5000 Actually, "Photoshop assisted" kinda already exists... so yes. Pretty sure we have AI assisted category for art as well, not to mention straight up a label for AI generated works that definitely does exist. "Thermomix assisted" for dishes...? That might exist under a different name as a category, somewhat. CGI for movies is almost a given these days, so really you are looking for what *isn't* CGI or something along those lines if there is anything -- it's also a thing where movies have basically almost grown up with CGI, with similar things already existing as special effects - which is a completely different thing when it comes to literature/writing as AI generation has come in much, much later in the game.
Yeah, we actually do care about tools that were used to create things. Partly because end results *are* affected by them, but also because ethics. Ceramics had a thing with 3D printed works being too fragile, so people stopped buying mass produced 3D printed wares since the quality sucked. There are people who have issue with resources sourced from conflict zones, such as certain rare earth metals for electronics. So yes, people boycott those and companies that use them as well. I don't think much more needs to be said about this
9:30 i totally agree; i prefer to avoid AI generated content when I can, especially as a writer myself. But i LOVE the authenticity of the AI in her story being based on a real AI's responses, just as authors will interview professionals or research how any professional field truly operates and functions to be more authentic; to me, thats preferable to just imagining what the AI character might say, and coming up with something innacurate or representing it poorly.
Kinda genius on her part, but maybe talking about doing all that in the afterword at the end of the book would have been the best place to mention it.
I feel like it's better to let the actual reception (reviews, rating, popularity, and where/how many ppl drop it) of a work (and synopsis + genre-tags ofc) decide whether I read something. Basically, Only thing I care about is that it is good, why should I care whether it was made using a hammer&nails vs electric drill&screws?
Honestly, this drama seems more like a case of people putting more value in authors (who avoid any "modern" tools like spell/grammar check or AI) vs proompters (that are additionally both authors and editors too, so it's not like she wasn't an author).
@@feha92 I imagine in a few years time, that'll probably be the case for the majority. But for now, people are wary and find AI ethically questionable; where/how is it pulling its data/responses from? How accurate are they? For artwork, did the original artist who spent however long making a piece consent to AI using said piece and essentially tracing it (+ however many others) to create a new one at the click of a button? Etc.
This won't last long I imagine; most people don't care about the child/slave labor in other countries enough to actually avoid purchasing or using anything made that way. Convenience and necessity outweigh that ethical conundrum, and I figure AI will be the same in due time.
@@netherslayer3561
"where/how is it pulling its data/responses from?"
From a machine learning algorithm, usually a neural network, that has been primed through a learning process and overall designed to produce nice responses to nice inputs. This learning process is ideally done through feeding the internet into it, or at least curated portions of it.
"How accurate are they?"
Arguably I prefer answering this with "not at all". Sure, the responses can be considered somewhat accurate in that the response often ends up being accurate. But overall that is just a coincidence that stems from there existing patterns in whatever in-out process it needs to emulate.
Effectively, things are as inaccurate as humans are, because both us and the algorithms we design to act like us, are merely pattern recognition/reproduction machines.
"For artwork, did the original artist who spent however long making a piece consent to AI using said piece and essentially tracing it (+ however many others) to create a new one at the click of a button?"
AI does not just "trace" the inputs (not even art-ai's like dall-e). Just like a human does, it 'finds' patterns in the process of "prompt for an image" -> "resulting image". Unless poorly designed and trained, it does not copypaste or trace parts of images - just like humans generally don't do that (though arguably that is more because we have so faulty memories, otherwise I imagine we would be way more prone to "tracing" things we have seen from memory)
Hopefully the artist didn't feel the need to explicitly consent (explicitly choosing licenses is meh!), and instead the ai was trained on stuff released into the void to be seen by all and any eyes. If you put out an image publicly on the internet, you can't expect people to close their eyes when visiting your websites until they have read the license and whether it allows people (or ai's) to see the image and process it into your memory. Honestly, it's silly how many people think it is a horrible thing that their publicly released images accumulates as part of the experiences that is other humans lives, just because AI exists now (if they think it is horrible that an ai is trained on an image, they also think it is horrible that a human sees the image and isn't bonked in the head before it enters long-term memory).
I think there is a strong possibility ai will become very limited, simply because there's so many vested interests and money in gatekeeping content-creation. So legislation is very likely to be created.
And it isn't even the worst thing if it does - I just find most of the more common arguments for it to be utterly idiotic.
Don't ban AI because it "plagiarises" (rather, use ai to streamline the existing anti-plagiarism pipeline in the justice system) as that's not actually an inherent problem for it, ban or slow the adoptance of it because it can destabilize society if very rapidly one of our larger sectors (entertainment) suddenly ceases to employ as many people as it currently does (granted, I _personally_ don't think ai-development will go that quickly, and we still have decades - at the least - before it becomes more than a mere tool in an artists toolbox. But at least the argument has enough merit that the trends/statistics requires observation, simply because the _consequences_ would be grave enough).
Don't ban AI because it outperforms humans while still being worse than the really good humans. Ban it because the same performance advantage hurts future ai-development as future ai starts being trained on the results of past ai (rather, just enforce proper and standardized tagging of things).
And so on.
@@feha92 Sorry, copyrights exist for a reason. AI generated artworks who use data scrapped off the web are violating numerous copyrights. There are people who get in trouble for referencing someone else's works too heavily. There are people who get in trouble with tracing (or heavily referencing) parts of other people's works as well. AI is getting a ban in the art world because it is trampling on existing rules, just as other artist get in trouble for trampling those same rules. If we let AI get away with what it is doing, that is essentially playing favorites as if it were the "cool new thing" instead of properly enforcing rules and regulations. Why should we let a bot get away with things a human isn't even allowed to get away with? Letting AI get away with those violations makes very little sense, especially considering the damages from AI are considerably more severe and quickly repeated over and over again
The reason AI is not entering artists' toolboxs is because there were too many blatant copyright infringements done in the development of AI art generators that numerous people don't want to get near it. *IF* AI art generators did not have that phase, they would have been under consideration as an artist's tool; however, the reputation of AI art generators has been permanently and irrevocably stained due to the careless actions of several AI art developers. Because of these copyright violations, generated AI art is ineligible for copyright in numerous countries. This would not have been an issue if AI art generators were developed ethically by getting proper permissions to use the artworks they were trained on. There were other methods for creating AI art generators as well, with actually properly studying an amount of art theory then converting those into basic formulas to start the AI art generator development. And what'd you know... so much of art has already been boiled down to nice and convenient basic formulas for these people to use as a base -- except they used NONE of it
If you want to hate how AI is getting blocked in development legally, you can hate the developers who decided to break the rules in the first place as well as the users who decided to violate ethics and utilize the tools for nefarious reasons. You wouldn't be seeing these issues if they actually followed proper procedures and considerations
@@gaerekxenos Copyright laws exist for a reason, and ai being trained on them violates none of those laws. I have yet to see an artist get sued because in their childhood they saw a drawing or movie by disney or nintendo (or even by random ppl), and we have had centuries where both artists and copyright law has existed at same time.
What you mention is when said influence is _too_ heavy (ie. tracing), and it is up to the individual to not start publishing art that violates it. The ethnicity of the artist doesn't matter, it is the final result that does.
AI is not getting a ban from trampling on rules (it doesn't do that), it is getting a ban because people are afraid they will lose their jobs/purpose, and grasp at any straws they can to stop it.
You should always reframe the discussion about art-generating AI by pretending you are talking about immigrants from ie. ireland (that was the american stereotype, right?). Say a couple comes to your town and has kids. The kid saw your art growing up (alongside all other pieces of art in the world) and decided to enter same field as you. They turn out to be really prolific while taking less than minimum wage, putting you out of a job as you simply can't compete (they are probably doing a better job too!). But you can't take them to court over copyright issues as none of their art actually looks like yours.
So you claim copyright issues because you know the school in your town had a drawing you donated hanging in the entrance, that they must have seen as a kid.
Hi Joey, I just wanted to say that I love those videos where you cover Japanese news. Please keep doing that! ❤
Keep up the awesome work and videos Joey love the video
Yeah, I think some transparency is gonna need to be put in place. Both for the organization giving the award(s) and the participants putting in their works.
Imo, the best thing would be for the participants to note and describe the use of AI, if any. The organization will probably need to come up with a list of banned methods of AI use along with a pre-approved/commonly approved uses of AI. That'll probably take time though unless they can come up with a way to speed up that process
Don't ask don't tell. I'm kidding -- that's a terrible joke.
In all honesty, this particular case where she used AI to generate an AI character's response makes sense and is a pretty special scenario. Typical AI story generation from what I have heard tend to generate bland and generic tales. The smaller snippet honestly sounds like fair game considering the context of its use. Obviously, going forward, there are going to be some issues regarding AI regardless of current controversies -- assuming it develops in terms of quality and they keep feeding good quality works into it
I do agree that the best thing to do would be for participants to note and describe the use of AI as part of their submission form. They don't have to outright ban methods of AI use -- it just needs to be noted so they can make appropriate considerations regarding whether the use should fly or sink. In my honest opinion, I don't think they were ever considering AI generation for text as it simply... doesn't work that well in terms of story writing. The current logic for AI is considerably lacking, so generating an extensive piece of literature that would be relevant to society and culture simply isn't feasible. Using AI simply isn't and wasn't on their radar for literature
In all honesty, it's almost a bit of a "go ahead and try" sort of deal for using AI for literature. Do you really want to go through hundreds of pages of text just to check whether or not something is a complete and comprehensive story? Multiply that by a large number attempts with the AI generator? I sure hope you enjoy reading. LMAO. Image generation is considerably different -- you have only one piece to review, and you can do that review very quickly and almost instantly. That is not the case with text, aside from reading the first sentence and immediately seeing that the entire piece is complete garbage and discarding it because the start is complete nonsense. You have to go through the entire thing just to confirm something is passable or would work as a base. It honestly sounds like nightmare fuel
Well i understand what you're saying. Waiting until the worst case scenario or when we are getting to the worst case scenario is a kinda dumb. Whe should at least be setting up safety nets or something.
The only thing I dislike about AI is using it not as a tool but as profit. You can't just use information taken from AI and that's it, you need to elaborate it.
Same thing with art, you can't just sell an art made by AI, use those details given by AI and make something with your own.
I had to take a writing class in college where I had to write multiple papers. I used ChatGPT to give me ideas on bullet points and layouts for a specific topic. It helped immensely.
That’s cheating!
@@paulwalther5237How is that any different from getting advice from a teacher or student?
@@upg5147 The computer is smarter.
If you're using ai to help you write ✍️ a book 📖 then what's the point if you're not writing the book completely by yourself
if you’re using calculator to help with a math problem, then what’s the point if you’re not doing it completely yourself.
See the problem? People keep thinking that AI is doing all the work when in fact it is just there to assist with work.
Think of it like a calculator, it’s there to help you with things. AI is used to give you ideas, tell it to suggests a bunch of settings and when you find one that you find pretty interesting, you start working on it.
At the end of the day, everything is still up to the user how they want to use AI.
@thisisnotaviableaccount it's true but don't forget the more you teach Ai the more it learns but I never said that Ai can do everything by itself
One artist youtuber brough put that there's highly AI-generated art tutorial books being sold. It's whole another situation than author using AI tools for how AI would work and respond.
My mom works in the US public school system. They quite literally just had an inservice day where they were told that there was absolutely nothing wrong with students using AI to complete their writing assignments. Also, apparently it is no longer okay to use the term plagiarism.....you have to call it "academic dishonesty". I mean, WTF?
I think it would have been cool if she had mentioned at the beginning of the book that the AI responses are real AI. Like a book that starts off with "based on a true story". That way it also wouldn't have seemed like she used it to "cheat"
06:55 Unsurprisingly, the ones who make the loudest noises in the media are those who have never read the book or understood the context of the author's use of AI.
Refining your work with ai is fine, creating your work with ai is wrong imo
How would you define "creating your work with ai"?
@@Kyouma.Just utter plagiarism
@@dunzek943 that's not how AI (at least not LLMs) work. And if it had worked like that, it would have been useless and noone would have cared to use it.
But don't worry, plagiarism already has systems in place to be kept in check, so AI won't affect anything there. Just like you can sue for plagiarism done by a human that wasn't ai-assisted, you are also able to sue the writers that _are._ Whether they metaphorically used a hammer or an electric drill to build their table, the law treats it the same - as it should. The tool doesn't matter.
People may gush over AI, but I could never use it to write a story. If it doesn't come from my own imagination, then what's the point? If all books were written by AI, I'd give up reading entirely.
5% is negligible especially when she wrote a book about generative AI using a bit of generative AI, kinda brilliant also you can find out if something uses AI, better admit now.
As an author I fucking hate AI and chatgpt but eh...
Any news on that crowdfunding thing, my man?
"AI affecting author creativity for LN contests"
My brother in Christ, most contests are the same copypasta creatively bankrupt submissions that it may as well be AI generated.
Japan's business culture of "Only publish/produce what already works" so most stories coming out of these contests. AI will do a better job of spewing out the isekai/LN funnel.
Mind you. You will still get the every so often Original Works that actually breaks through, but by in large the market is flush with blatantly empty fast-food writing will just copy and rehash the good works just like an AI would.
Gen AI that is incredibly unethical is that it rips work from creators who never consented to be a part of it and then is used to force them out of work as they get replaced my a machine. This is already happening and will keep happening until the laws catch up, but by then it will be too late for many creators.
Gen AI as it is today is morally reprehensible and anti art, and I wish more people could understand that. If the consent model was opt in rather than opt out (which is impossible to manage if you don't even know an AI company exists - moreover it's ridiculous that the burden is put on the people affected rather than the huge company trying to profit from them), that would be a different story. But the big AI companies know they can't do that, because too few people would opt in. Additionally, there are sensitive materials in AI databases that break privacy laws such as medical records and other things that are much, much worse.
Some people argue that it democratizes art, but the truth as it is, is that it's the absolute opposite. It rips artist's means of living away from them when they've spent years and years honing their skills and getting to the point where they can get hired for it. And because of this at the moment, even this kind of use of AI is honestly not ok. Do I agree it was clever? Could it have been cool? Sure. But doing this sets a very bad example when gen AI isn't ethical yet.
It’s not going to be either extreme. AI is just going to be used by corporations as a tool that lets them pay us less for worse jobs that produce shittier products but make shareholders more money
2:50 Atsushi: *AKUTAGAWAAAAAA!!!!!*
Remember folks. We live in a time where almost any information is in the palm of your hand, but actual intelligence has hit an all time low amongst the general population.
First the Hugo’s and now this, not been a good year for literary awards
Edit: okay that specific case of using ChatGPT to show what ChatGPT sounds like is probably the only way use of ai would be acceptable in my eyes. In general though, my favorite way I’ve heard it said is: ai is a plagiarism launderer. In most cases the use of ai generated text for a published work is blatantly unethical. Hats off to the author for finding the exception to the rule.
I think that is cool -if true- about asking a question as the characters and getting a true AI response.
how would we feel about a story that involves AI characters (culture) and the writer asking an AI what it would do?
Fitting those answers into the story(world) to me would be interesting.
I actually don't think this is a big deal. I think it's a bit like "Memoirs of a Geisha", where the author actually spent time with and interviewed an actual geisha to make sure that the experiences of the characters and stuff was all realistic.
Honestly I’m more worried that big corporations will replace workers with AI whenever it’s possible because it’s cheaper, leaving many people jobless or locked out of the careers they want. It’s only when money/work gets involved that AI is scary lol, her use of it was perfectly fine I think
In regards to writing & illustrative art, AI's database has been trained with stolen works and no compensation whatsoever to the artists who now have just become style prompts in this database. No matter how "creative" you think you are with your prompts, it's not something you really created.
It was not created to aid artists as a tool (It would be cool if it was created as a tool aid, but it' not the case), the same creators said that making art is a hassle they wanted to skip, they did this so that them and other companies don't have to pay artists for their hard learned skills. Now they are spitting out images like a randomize collage maker and people are patting themselves on the back for typing a few words on the PC and calling it a day.
I am not sure how the law is in japan in regards to AI u se yet, but here's a piece of the European AI act
"Text and data mining techniques may be protected by copyright and related rights. Any use of copyright and protected content requires the authorization of the rightholder concerned unless relevant copyright exceptions and limitations apply".
even if it was jut 5% that was used on her book, that 5% is mashup work stolen from who knows where.
Honestly, she cheated. If you didn't create something on your own and have to use AI, it's not fair. I can understand using reference materials like actual books and historical literature and texts, but she literally used another 'brain' that has every knowledge in the world, and its lazy. It's only fair if others get to use AI too as part of this contest.
I hate to bring religion into this, but in my church, we have bi-annual broadcasts that are 10 hours long, 6 hours of spiritual talks on a Saturday, and 4 more hours the following Sunday.
The broadcasts are originally in English, but are then translated into other languages throughout the world. However, due to not being able to translate all the talks, some countries would not receive the translations due to the talks still be translated.
Thanks to AI, the Church has been able to immediately translate all the talks into dozens of languages in a single day, and I do assume that there were some errors, but a quick editing could fix those errors before they the broadcasts.
Another thing. Journal keeping is important. It helps you connect with your distant descendants. But one day, this Sister in my Church said how her great grandchildren won't be able to read her journals because it's all in cursive.
AI has been able to help her with that by transcribing everything she wrote into common writing.
So yeah.
AI, go do your thing (responsively).
Interesting video I’m actually studying generative ai and one of the books I’m reading about the topic was made by chat gpt. the author used to be on the board of directors of open ai and it has a lot of good reviews. Although he did preface that chat gpt was used at the beginning of the book
I think that we're far far faar away from any of these doomsday/"judgement day" scenarios. In fact, using the term AI as an umbrella for machine learning, deep learning, etc, what actually has progressed in the real world, is a misnomer because it is infantile as compared to its AI movie counterparts. The latter is technically known as General AI, when the machine begins to think for itself, like Terminators. ChatGPT? It merely has an absurd amount of training data available and uses probability based algorithms to generate human like answers.
Our AI is in its infancy and has no emotion to write books, movies, TV shows or comic books.
ngl I got instantly reminded of 'domestic girlfriend'
what? context plz
How? Thats so random, lol
@@ryana5435the protagonist of that manga is an author
Lol this guy didn't even read past the headline before hitting record for the video.
He typically does that, it's called a fresh reaction or a first impression.
yeah. 6:57 "Kudan explained later in an interview that she only employed Al-generated text in the responses given by Al-built in the story." so we got clickbated by this video's title, and Joey is clickbaited by the news article's title.
@@ryana5435 Ye, exactly what I thought. Man could at least read the article before trying to explain it to an international audience.
@@ryana5435 clickbaitception, we really do live in a society
Okay, so I am not crazy for finding it a bit "funny"
I dont think it's good. I dont know what to think about creating ideas either. If i struggle to create a story and ask chat gpt to make examples as I wished them to be, it's no longer my creativity, that part is not my creation. To me it feels like cheating, anwsering questions with use of AI. It's just a different form of taking a test, or an exam, instead nobody is going to check if you are cheating.
We do get the option to choose what piece of Japanese fiction we want to read later this semester. Maybe I should suggest that book.
As an aspiring novelist, I can see ways AI can be helpful without being fraudulent or morally questionable.
I can see AI being very useful as an editing tool. For those that don't know, professional editors can be quite expensive (as in the 4 digit range for a single edit of one book) AI editing tools can help writers who can't afford to hire an editor. Though I think it can replace a developmental editor, an editor that helps with the foundations of a story.
I'm not sure about writers using AI tools to change the voice of a book. Writers have to develop their writing voice on their own. If they have a app do it for them, then it really isn't them.
I do not agree whatsoever with using AI to completely construct a story. Maybe it could be used as a tool for a writer to bounce their ideas off of. But not as something that you just feed a vague premise into and just take what it spits out. There's no humanity, passion or hard word involved in that. And humanity, passion and hard work is the fuel for any creative endevour.
So my feelings on AI is complicated. I think it can and maybe should be used as an assistive tool. I don't agree with the AI doing all the work and a the writer claiming it as their own.
Joey it already has happened, someone used AI to write a whole book and a AI generator to do the illustrations and made money of it, he put 0 effort into creating it
I think the author saying they used AI for the novel was a genius marketing move! Everyone is now talking about her.
Her revelation about AI use makes perfect sense to me. Using AI to write a story about a world of common generative AI use gets "closer to the metal". Nothing gets closer to simulating generative AI behavior than actuallu using generative AI. She's doing humanity a service by forcing us to face the reality of how powerful AI can potentially get.
AI wave is like the smartphones wave, either we get smarter or we get dumber by it,
While AI can be used as a tool in creative works, we should also thing about the inevitable reality fully unregulated use of it will lead.
It would be super funny in a future where AI gets good enough at story writing (and realistically it's not that far down the road) where it wins an award and the person behind the AI just imputed the prompts and nothing else. By all accounts, people will likely just make different categories to separate contestants based on how much AI was used (like 0%, 5%-10%, etc). Though of course it will be practically impossible at a certain point to know if people who claimed to use no or little ai, just used it for the whole thing. Unless we have a smarter ai that can tell.
Not to mention in terms of sales, knowing how lazy and uninterested most human consumers are, a lot wont care if works are generated on mass by like a single dude who imputed different prompts for what ever genre of books they want. While it may take an author months or years to finalize a version of their work that they are happy publishing, it will likely just get drowned out by the millions of books auto-generated between the start and end of their work process.
While people may create works of literature for the fun of it and AI wont take that away on a personal level, in terms of competitions or financial gains on a wider scale, creative human writings will just be outclassed by superior AI works.
Generative AI a bit hard to use what you want it to create, I try to generate image for styleframe still take time to create a prompt correctly
nice, so she used AI to generate the responses made by the "AI" character in her book. makes the ai in her book more authentic i guess lol
The way she used it is unique to AI programs and thus entirely fine but I think using AI to get the ball rolling or editing work is entirely fine so long as the human does a majority of the work. I see it as no different than hiring an editor which any good book should and usually does.
It would be great if this story become anime
For the context of how she used AI, i think its okay. but I dont think its okay for people to use the AI to do the work for them and being rewarded for it. specifically in work, competitions and College. You should absolutely not be using AI to help write your thesis/dissertation. That is absolutely cheating. also there is no Copyright protections anything producted by AI. although the bigger issue it where the AI pulled that information from to produce AI work, the information is definitely copyrighted work. The AI doesnt make up stuff without using copyrighted material. it all comes from somewhere.
I have used ChatGPT to edit wording/grammar/tone in emails, job application materials, etc... In my workflow, it functions as an incredibly fast and proficient copy editor. However, I have to either hand it nearly finished material and or heavily edit its responses when it comes to domain specific language. It’s a tool and nothing to get bent out of shape over. I think this author’s use of generative AI in their book is a neat application! Most discussions about generative AI right now are just fear mongering.
Fear mongering is bs, man.
I’m not anti-AI, but I think using it for a competition where the point is human creativity, is super uncool. I think this lady should have her aware stripped.
She just made her book a bestseller by admitting it
only %5?? did she asked "Correct my grammar and spelling mistakes?" or something ?
As I writer myself I can never EVER imagine myself ever using AI. Using AI implies a massive contempt for writing as if its a task to be done instead of a labour of love and expressive creativity.
I would rightly look in contempt towards a person who would use AI in their writing. It simply means they don't have the chops. Its a highly competitive field and people are very proud of all their hard work in getting good at it. Me included. If someone just used AI to do the heavy lifting and did a bit of light editing to be legally distinct enough to avoid a copyright lawsuit. I wouldn't wish these people any success.
In prose there is far too much intentionality to be left for AI to create. From word choice to sentence structure, none of it can be trusted with AI if you're a writer worth a damn.
I think you are too far on one end. I agree that you should not let AI write a whole story and then edit it a bit to be legally distinct but the opposite is fine. I see that as no different than hiring an editor.
From one writer to another, I couldn't agree more.
Personally, I feel she should've admitted to the use of AI when submitting her work cuz it seems pretty unfair that the contestants who wrote stories of similar or equal quality to hers without the use of AI got shafted out of winning the award.
meanwhile we don't need A.I. to produce a crap ton of Isekai light novels because the authors are already just flesh and blood "A.I.".
This IS a publicity stunt.
Using currently existing AI tools to produce the responses from her AI "character" is actually ingenious. Just as ChatGPT cannot currently perfectly replicate a human's though process, neither can a human perfectly recreate an AI's algorithmic "thought process." I believe, however, that prior disclosure of her use of AI would have resulted in the disqualification of her novel.
I'm not surprised. My students at the uni use it all the time and sometimes it's so obvious that I just soft key fail them with whatever excuse that works (it tends to be the students who never work at all, are always late with reports, etc, so they know that we professors have more than enough ammo to justify their F), and they're so stupid that they don't realize that maybe even if they man age to fool us, if they spend 4 years in college learning jack shit, they'll probably be sacked at their jobs after 3 months if they even manage to get hired. I mean, I know that Japanese students learn the very spcific and limiting job they'll perform for the rest of their work life at a big corporation in the first 6 months of working there and that whatever they studied in college matters jack shit in the end, but if they're already struggling now, imagine after 4 years of having learned NOTHING, trying to learn to do a job or be fired. Ah, time puts people in their right place, and I'm sure we'll start to see an absolute catastrophe in the workplace as a whole in Japan in under 4 years.
Yeah, if they use it poorly so they fail the assignment, you should fail them. It is only when they use it properly and prove that they understand both the assignment and the tools used to solve it, that they should pass. The same is true no matter what tool they used, if in art class you ask for a drawing of an apple and someone chooses to use watercolor but fails to draw an apple well enough - well, they failed.
Just like there is no inherent issue with people using LLMs to solve assignments (unless the assignment is fundamentally flawed), there is also no reason to turn a blind eye to their mistakes stemming from them using the answer verbatim without properly knowing the topic and fixing the flawed answers.
Either way, they will fail the exams if their use of AI was poor enough that they didn't learn (assignments are there to help you learn the topics of the class, old exams complement that, and then the real exam is used to prove that you succeeded in doing so) - since they either can't access it at all there, or lose time due to pens and paper lacking a copy-paste functionality.
I dont even trust my older brother to wash dishes correctly, and people who created these AI programs actually thinks people going to not misuses the hell out em?
Sigh
Just because some people misuse it, everyone should stop using them? People use knifes to kill others, but no one suggesting forbid them
@@darunggar6349
>some
Mate, everyone misuse it
I never once thought the whole AI thing is a good idea since people doesnt have morals and self-restrain
I heard another one of a guy using AI in a photograph competition just to see if he could and he won first place
Nah she is bullshitting. Take her prize money and award away. This isn't fai to someone who didn't do this.
It’s brilliant marketing!
Well I'm glad she was honest but man people are going to be doing this and not telling the truth Anyways thanks for telling us Joey
5% is about as much as someone would use Grammarly or autocorrect in MS Word. Non-issue.
Microsoft Word's Editor, Grammarly, and any other technology based tool modern day Writer's use-is already using generative AI. ChatGPT and other generative AI, are just more advanced. EDIT: also, the easiest way to battle the IP thing. Treat it like business as usual. If a entity or person, finds out another person copied their IP (to the point it is obvious, for example I can't believe Nintendo hasn't taken PalWorld to court), just let them sue and take it to court.
I'm guessing my comment got dusted lmao
Joey you should cover the Yui Ishikawa situation cuz these people are weirdos. The guy that got sentenced for threating Yui Ishikawa, I'm sure you can find it pretty easily.
The purpose of the Wikipedia rule is that they want you to not just pick the first thing that comes up on Google when researching. This rule is frustrating because they don't say that outright, instead teachers and professors often slander Wikipedia as an inaccurate source even though most pages on the site are vetted daily by both Wikipedia staff and users.
You'd be suprised at the number of people who admit their wrongdoings or at least questionable practices. I've seen it a lot during my school days and working life. These people simply don't see what the problem was in the first place, whether it was indeed the wrong thing to do or not. It's really a matter of needing to remember to put yourself in other people's shoes and look at things from their perspective outside your own, because often times it _just isn't a good look_ when you change your POV.
Eh, based on her description i dont see an issue. It seems she still wrote the vast majority of the book herself and used ai primarily for a specific ai character, which while gimmicky, isn't the same as just writing a whole book with ai.
It depends, really. AI has been used as a tool for a while now in design/animation industries, but there's a line between being used as a *tool* to aid actual artists (like in this case where the writer asked AI questions like an interview for authentic answers) and designers VS *plagiarizing and replacing people*. I still am not into AI "art" mainly because there are no laws that dictate that people should be allowed an out from having their art fed into AI without their permission. I got threatened by a troll with having my piece be copied and I do commissions without using AI, and plagiarism's also a huge problem.
Hopefully we get the field to progress far enough soon, that AI is less of a tool for an author to use, and more a tool for a proompter to use (no actual writing or even editing needed). For now, any good work that used AI will only ever have had it been a tool, but in the future we might get it to write full stories on demand as desired (preferably even readers of other works being able to submit those and ask for "works like this fav of mine" or "continue this story", or fanfiction).
@@feha92 I still hate it even after seeing the nuance for its use. It got to the point that people get rejected for jobs because of this.
Ai is great to buy tame aka 手間を省く so it makes easier to to the research and get the necesarry info quick to you finger tips
AI can only learn by stealing from human creators; always without their permission. Using it at all, even for 'inspiration' is stealing.
Did you read the book?
Okay so more like documenting the response of the AI Okay got it 👍
Say if I’m writing down a story that I made, but I’m not a certain if it’s not good or not I simply use AI just to put improve it and added a feature, for example “improve it, but make it British” and it did and it sounds good and ok. But you could add a little detail on your own or edit it. But to me, I think it could be used as a co-writer
7:09 Joey changes his shoes on the go Lmao
i dont see any issue with that usage of AI in this case.
The game is the game
Nothing wrong with working with machine intelligence to create works of art as long as the machine intelligence is credited; the real issue is, she took all the credit for herself imo. That is what bugs me...
Has anyone asked AI what do if there is no electricity to run internet infrastructure?
AI use in science and medicine is 100% Ok , but on the arts is lame and shit and a crutch , art is the only thing that really make us human and surrendering it to Algorithms is ashame . Also calling it an intelligence is a disnomer ... its just effin algorithms it should be called algorithmic generation NOT AI!!!! And yes we should be told if something is made with "AI" so I wont support it at all.
as much as artist and others want ai to go away it is here to stay. we must adapt to it and help regulate it rather than scream that it is a vast evil
She just wanted to get more eyes on this. No one would know about this if she didin't say it.
100% I think this would've been fine if there was a note stating that some AI was used. Even in the fine print of the credits at the beginning that people skip over. Or even at the end of the book. Some time before the award was given out, at least. She was probably trying to make a message about AI with her book, a message that would've been way more genuine and powerful if she was, at any point, transparent about her use of AI. The fact that she hid it from the print is almost admitting she thinks what she did was unethical.
If people choose not to buy her book because of AI, that's their choice they should be allowed to make. I think that'd be better than backlash after the fact. I appreciate when AI generated image posters (so-called "AI artists") are transparent about their sources because I can block them immediately on social media. They're minding their own business, though. I'm just livid when people aren't upfront about typing in a prompt to spit out visuals taken from, y'know, Shutterstock or private health records or wherever else they've proven to have sourced content without consent nowadays...
Surprising how you reacted emotionally, until you've realised how it was used. Also, you've said that nobody would've found out; I guess someone has and they might have threatened to leak it so she said it herself first
This reminds me of Socrates bitching about writers using punctuation, or academics griping about using calculators because it will destroy humanity. People are dumb.
Consumers don't care about how, but what they get. Tools are irrelevant if the product is good.
She should add ChatGPT as a co-author of the book.
Since it was for 5% of the book and less than a full page when added together of what I assume is well over 200 pages, I see no need. It's like asking a friend for one or two lines or an idea you are stuck on and saying they need to be credited. If anything, that's a special thanks.
@@upg5147 Good point. She should definitely add a special thanks to ChatGPT then.