The Internet is Fake (Dead Internet Theory)
Вставка
- Опубліковано 26 чер 2023
- The internet is ever changing, but over the past couple of years things have started to change. The proportion of bots in increasing. A theory dubbed “The Dead Internet Theory” that states that most of the internet is in fact AI bots. These bots have been unleashed to control the populous and rake in profits. But is it true? And if so, how much of it is true? In this episode we’ll find out.
ColdFusion Podcast:
/ @throughtheweb
Outro Track: • Burn Water - Eunoia (N...
ColdFusion/ Burn Water Music:@ColdFusionmusic
ColdFusion Socials:
/ discord
/ coldfusiontv
/ coldfusion_tv
/ coldfusiontv
Producer: Dagogo Altraide
Writer: Kate Rostance / @ThenWhatHappens - Наука та технологія
![ChatGPT Can Now Talk Like a Human [Latest Updates]](http://i.ytimg.com/vi/ID5tc61iksY/mqdefault.jpg)
Just some clarification: I'm seeing a lot of comments talking about how I said in 2021 "there was no widespread generative AI". By that I didn't mean generative AI didn't exist before 2021 (I've been doing videos on them since 2018) What I meant was that it wasn't "widespread"
Just wanted to clear that up. Cheers
WEF could be that group controlling the net
Maybe those people were actually bots?
All that time and energy we waste arguing with bots can't be helping society 😵💫. We need a new social media platform that requires a heartbeat to sign in.
Wait when you said “no widespread generative AI” you meant no widespread generative AI? That’s crazy 😂
@@Sum1Better lol good joke
What a cool dystopian short film it would make if a guy who spends all his time cooped up in his apartment on the internet finally notices he's the only person alive.
That could have something to it.
Black mirror
lol I could see that working out. Like it's to the point where everything is automated so there's still power and water and whatnot because all those systems are run by AI. And he gets his groceries from delivery services that use drones to bring them to his door. But over time he notices that every time he goes to pick up his groceries the world is getting quieter and more and more overgrown.
🤔🤔 this is actually a good idea..
Post-credits scene/sequel: He realizes not even he is alive...anymore. Everyone is dead and are just AI avatars of their past selves.
The worst part about this is that it has barely been 10 years of the modern era of internet (post MySpace/askJeeves) and already the internet is becoming unusable due to bot accounts and bot spam, imagine what this looks like in 50 years lmao
It's sad because it is such a stupid and fixable problem. But our corporate overlords have no incentive to fix it and our political class takes cheap bribes to maintain the status quo.
A metaverse full of 'idiocracy' AI?
IRL they can't even change the red light for my fucken semi.
i can't.
Without adblock the internet already looks extremely insufferable. Can't wait all the new ways our corporate overlords are going to come up with to force their ads on us.
@@littleponygirl666neuralink straight-up making you want to buy certain products subconciously without showing ads
"We could be entering a post-truth world." I want to ask: how the hell are we not already in a post-truth world?
Worth asking if we ever lived in a world of 'truth.'
@@ZPS51491 Good question.
True sense the beginning of intellect reality has been seen as an illusion of sort
I'd like to think that the Internet becoming so untrustworthy will one day push people out of the digital realm and enjoy outside/real-life more, but I highly doubt it...
to be honest i dont think people will care if who they are dealing with is AI or human as long as they get what they want from it.I mean if was looking for a history documentary channel and found one without a human presenter just a voice but it was well done and educational I would subscribe even though its AI
AI is going to make it ever increasingly difficult to discern truths from mistruths, and humans from bots. It's unavoidable. Even AI designed to identify AI won't be able to tell the difference. At some point (probably not that distant from now) only in-person human interactions will guarantee the latter half of the problem-that one is actually interacting with a human. (We can barely trust videos anymore thanks to incredibly convincing deepfakes.) The "am I being told the truth?" part is obviously as difficult as it's always been, but it's greatly muddied by a sea of nonsense, lies, propaganda, and the like. Who has the resources to do the necessary research, much less the time.
Without sound foundations of truth democracy breaks down. It certainly seems democracy is in its all-time weakest state. The part that many people seem to miss, I believe, is the following:
Citizens of democracies would do well to be mindful of their greed, especially policymakers (those in power) because they are going to be the ones ultimately responsible when it all comes crashing down. And most of them will suffer the fallout just the same as the proletariat most of them furtively deride. So many policies are driven solely by making money rather than actually helping mankind.
lol it's already happening, you just don't know. people with hobbies like skateboarding exist outside bruh
2020 taught us. In the digital realms pants are optional. I pity the fool that gets the ubre after my naked ass has been up and down the seats.🙃
I live in an ex-communist apartment block in Eastern Europe next to the sea, largely I am here because this is still relatively real. Certainly way WAY more real than life in the UK or America. My neighbors sit on the porch of the building having beers, the kids play with the stray cats and climb trees. I watch youtube but have no social media accounts.
It reminds me of my childhood in the 1970's in Britain.
I've noticed google getting progressively less helpful and all articles are either an ad for something, or so poorly written it might actually be written by a dumb human or dumb bot. Either way, it has become useless to search for things anymore
thank you for this. I've been saying this and I realize people just think I'm crazy ...
Yea, now you’re required to add the word “Reddit” at the end of the sentence your searching about to get any real useful and quick answers
@@paranoidandroid4817 LOL. This is so true. When I was building my computer, anytime I would Google something slightly technical, I only found the answer I needed if I appended "Reddit" to the end of the search. I'm not even exaggerating this.
I noticed a big change between 2011 and 2014 boolean an key word searches were not working, when I would type video I would get a list of 20 different video sharing sites, home pages with videos and so on, now ßlmost everything points to Google products snd services.
Indian AI seems to be writing the titles of a lot of videos now ... impressive, and catchy. 🙄
Disclaimer: No AI was used in the making of this video. Thanks for watching!
That's exactly an AI would say!
Prove it!
Cannot verify this. lol!
To confirm that you are not a bot please type this emoji: ✅
you're not a bot, Dagogo - or are you?
Back in the 90s, there were over 100 player at the same time, real-time arcade games on PC, and there was not a single bot.
The bots are certainly with us. A few years ago, as a self-published author, I was trying to establish an online presence and opened a Google Plus account (remember them?). Every time I checked my account I would be so excited to find that I had so many new followers, but when I finally checked them I found that the vast majority of them had strange names and profile pictures of busty, young blondes. I had to spend a great deal of time purging them all.
I have a similar experience with SoundCloud. Virtually all the responses I get are promo spam.
Is an online presence important for an author? I'm working on a novel myself and I'd appreciate the input.
@@miguelcondadoolivar5149 it’s the common opinion that it is necessary even for a published author. Publishers do tend to look at your public profile when they’re considering your book. You’ll find lots of information online and examples of author websites and social media to guide you. However, one word of advice. Don’t let building a public profile distract you from your writing. Good luck.
They are not bots are they. Thanks for ruining my day.
Currently happening with my blog. I was so excited before, now I don’t care when I get a notification 😂
The internet has been “dead” for me more and more over the last few years, in that using it becomes less and less helpful/fulfilling. There are a couple of highlights still (iNaturalist, Wikipedia, some UA-cam Channels and Reddit communities). But over all, spending time on the internet has become more and more of a pain, source of frustration, and waste of time.
Even though I was hugely excited about the future of the internet a few years ago, I’m withdrawing from it more and more now, and try to limit the impact it has on my life as much as possible.
Wow same here. I've actually been cutting out the internet as of late and not because I purposefully chose to. It's just not as useful as it used to be.
Wikioedia is a corrupt corporation. I wouldn't trust their information. You can pay to have an article "edited".
Even when I go to a genuine website for a genuine product from a genuine manufacturer, the information is often junk. Companies no longer want to inform their customers about their products, they just want to bamboozle them with marketing BS. Just try finding out basic specs for a car for instance - after five minutes of clicking though every nook and cranny of the website - "pre-orders for the 2023 model are currently open" 🙄
When I enter a site and the pop-ups and ad fills start, I start cussing and closing everything. Do the people who create these sites ever actually click through as a user to audit the experience? Some are just unusable torture chambers.
@@Screw_This worst are sites that demand you to create an account otherwise the popup to login/sign up is stuck on your screen.
In academia, it is understood there are very few people producing content on internet and a massive amount consuming content with no interaction
Is that your field?
Ok now bully me
The current situation with reddit has made me think about that.
Reddit is removing third party apps which while make up a small percent of users those who seek out third party apps are more likely to be active&interacting users therefore removing third party apps would have a disproportionate effect on the entire platform
I agree. 99.9% of the time I just watch, rarely do I comment.
Who else missed aol, compuserve,prodigy or GEnie?
Dagogo Altraide, you have opened my eyes to a whole new vision of the world and beliefs. I remember how the internet was so good, and I gained a lot of experience that shaped who I am right now. I recall the website where I spent almost all my day, the forums and topics I followed, as well as the online games I played from 2005 to 2011. Everything changed after this. Facebook took over, and little by little, we all drowned in the social media nonsense.
I appreciate you sharing with us bro...
But I hate to break it to you, the entertainment and joy you've been getting from the internet before? They were at the expense of many more people's enjoyments in life, not only that but entertainment as a whole saps productivity and money from The general public(especially the low and middle classes).
I don't mean nothing by it Friend, I just thought I'd get you some more intel.
From your comment you seem like understand you the truth somewhat.
و الله أعلم.
@@manitro337 that an. Interesting point of view, can you explain more?
The person prioritizes "productivity" over fun.
I think that the real lesson here is that we should all go outside more, touch grass, interact with people IRL more and build real relationships. It’s hard, yup, especially compared to the convenience of tapping one finger on a phone’s display. But that’s where good mental health and a fulfilling life come from, whether we like it or not
I think in the future were going to swing back to spending more time outside, especially when it comes to Gen Z becoming parents
@@28_futabadon’t most gen z kids hate children now? Either because they just hate kids in general or “to save the planet from global warming” 🙄
I don’t see many of this new generation becoming parents. I see a lot of nursing home patients in the future.
All of our attention spans are gone even my mom's generation too. The internet has always been dodgy, I only used it years ago for filesharing. Untill the Google era after the dot com bust.
ColdFusion is the most convincing Bot on the interwebs …. Respect 🙏🏼
He's the real predecessor to doctor Noonien soong
In AI we trust
This is why we never Dagogo in any of his videos; he's just a bot. As bots and AI progress, he'll probably take on a form and start reporting himself. Until then, we have an interesting channel and a voice.
@@rabbit251 Amen, brother. Dagago is the bomb! Much love and respect to him …
@@rabbit251 You never what?
This theory was explored 20 years ago in Accelerando by Charles Stross. In the book a group of Ghosts spawned off by their owners trace to the intergalactic router on a super computer the size of a tin can via lightsail. The arrive at the internet router to find it’s original creators long since gone or dead, and the networks taken over by basically bot like Aliens and AI’s that will capture you and hold you in virtual worlds for profit generation. It was referred to as economics 2.0 and basically if fed off the populace that are apart of it. It’s pretty wild, and a great read.
This sounds fantastic. Just ordered a copy.
Your AI profiteering post is working. 1 ordered a copy and I'm reading the free sample.
Touche
@@matthewbutt2340 I have a hunch that Charles Stross doesn’t need the money.
@Flumphinator I'll take your word. For all I know this could be his only book and it's sold 2 copies.
Or maybe he's the Grisham of futuristic AI books?
@@matthewbutt2340 He writes a lot of different stuff. I particularly like his Laundry series. Its a modern spy thriller in a world with eldritch abominations and mathematical magic :D
Another great video. The prompt of a dead internet reminded me of another topic you may want to consider covering at some point - link rot across the internet
I think if there's more awareness of the overwhelming number of bots online it will push people back into their communities and building real relationships.
Which will be a positive for humanity moving forward.
The downsides of AI has become more apparent lately as some people had time to get comfortable with it and start taking them to the extreme. Increasing number of scams, captchas on every website that doesn't want to avoid the inevitable spam. Major platforms hosting user generated content tripping over themselves trying to lock down, take ownership of the content and train AIs on it.
Worst of all, I no longer want to write guides and upload open source code that I wrote myself, because I simply do not want to help train AIs for Reddit or Microsoft. I'm almost certain there will be a sharp decline in generosity of sharing content, including other fields like artists, while AIs start overfitting by training on other AI generated content.
if your goal was purely to help people then it would still be helpful for AI to learn it and share it with people. although i can understand that it sucks, if they would use AI for themselves or take money for it
@@rencosa7827 Helping people, not the companies behind the AI. By providing training material for the AI we are handing it a tiny bit more resources to displace us. I don't know about you, but a world where none of our actions have consequence seems pretty depressive to me, regardless of the level of abundance.
@@iverbrnstad791with AI, there will be a point when we don’t need most corporations
Anybody would be able to make their own apps, do their taxes, help them study, 3D print most things
Indie development is only gonna become bigger
@@noirekuroraigami2270 I mean, we're at the point where doing taxes is trivially simple, assuming a half decent govt(I had one session of 5 min like 18 months ago...), besides that, study for what?
If the AI can do any intellectual task better, then why study? you might be able to make apps, but who will use them? You might be able to 3d print anything, who will want it?
We're not close to obsoleting companies, they will have the most powerful AI(or small chance Stanford or Tsinghua), they are producing the chips, owning the web servers and supply chain, large portions of our infrastructure, etc.
We might get abundance, but do we get agency?
@@noirekuroraigami2270 You're daydreaming hard, unless you could, somehow, elaborate how "the point when we don't need corporations" might be achieved.
What’s scary is how difficult it was/is for most people to distinguish between bots and people BEFORE advancements in AI. With the advancements we’ve seen now in gen AI it seems pretty hopeless for there to be improvement in distinguishing between bot and person.
Most humans are evil. I am fune with bots on the Internet.
@@CordeliaWagnerwho hurt you? 😂
@@CordeliaWagnerno?
@@CordeliaWagnerI hope you find your happiness.
The majority of bot traffic online is not using AI or machine learning, they are simply computer programs that run in an automated way. These is a big misconception and cold fusion is misleading the public with a video like this. For example, anytime Expedia looks up airline prices on the web, that is considered “bot” traffic because a computer program is doing it automatically….
Chatgpt literally advised me that unless humanity rebels against this corrupt system then Ai will be used against us with the chance of affecting any meaningful change very unlikely ..even gave me a 10 point bullet essay on how to implement direct democracy and empower and engage those uneducated and alienated . It was pretty amazing stuff .
Ok Mr Bot!lol
That's just because it found such content written by humans. Currently AI is more like a child who repeats what adults said without questioning it
@@tomlxyz ofc but it was a superb analysis .
@@tomlxyz That’s wrong and true. Because they dumbed down GPT 3.5 and 4 so much. Quantization fucked it hard.
It just told you what you wanted to hear
Worth pointing out that every website costs money to keep online, so I think it's more likely we've reached a saturation point from millions of people trying to make money online (recipe blogs to scammers) than anything else. Generated content and automation/bots are just tools, not the underlying incentive
Yea its like I think the film industry is dead and there are no good movies anymore.But at age 49 ive seen so many movies in all genres that another spy thriller is never going to get me excited.Seen too much of the same thing repeatedly
If my dating app experiences were anything to go by then, i'd have to agree with this theory 😂
If my Twitter experience were anything to go by, I’d say the internet is closer to 100% bots.
Dating apps are cancer. 10 years ago they were utterly awful, I just can't imagine what they're like now. Even back then I wondered whether many of the so called matches were actually real.
Guess it'll be a bot apocalypse instead of zombies.
@@muhdiversity7409Last I used such an app was 2020 and my biggest issue was getting recommended dead accounts or accounts that didn't fit the parameters I set. I'm certainly quite picky, but I'd rather be told I have no matches than to spend time swiping through accounts that clearly haven't been updated in ages or don't fit what I want.
I agree, dating apps are filled with fake humans or scammers
Putting trust on Google and Microsoft on safeguarding AI: we are fucked.
GoogToob and GoogNoos. 🙄🤷♂️
the second i heard that I audibly sighed. time for a cabin in the woods i guess idk
what are your alternatives? the US govt and the Chinese govt? the UN sets up a special AI committee?
@@perfectallycromulentWe need as many governing bodies as possible to get on the same page with regard to the various risks posed by advanced AI (including extinction risk).
Is that possible? I don't know. But when I imagine existing in 20 years, it seems very likely to me that I would be in a timeline where that occurred.
In other words:
"It's not possible."
"No. It's necessary."
SkyNet already won, gg.
Using AI to seek out other AI sounds like a potential paperclip problem...
"Send out the Alexa, or we will kick down your door, human!"
One critique I have with this video, is that the Turing test is not an actual test, but rather a thought experiment.
AI doesn't have to be that sophisticated to manipulate humans though. For example, it just needs to post one triggering comment, then have other bots up-vote that comment so it's prominently featured, then humans will take it from there and argue with each other generating division. So if your goal was to create distance between political groups in an other country for example, then all it would have to do is slightly nudge them towards generating more and more anger towards each other. With a large enough bot army it would be extremely easy and wouldn't require anything nearly as sophisticated as GPT-4. The same can be done with bots that are only there to up-vote certain content and increase the view counter. It doesn't need to be sophisticated, it just needs to target content that serves it's agenda then making it look like a lot of humans are engaging with that content, because that will draw in more real humans. So if you wanted to for example sway humans towards a particular perspective on a conflict or whatever, you could just deploy an army of bots to seek out and interact with that content until it seems like that content represents the prevailing opinion of humans, and many real humans will be swayed simply by believing that this is the perspective of their peers, even if that is not true. So while GPT-4 and other sophisticated AI certainly has the ability to make this problem worse, it's already been bad for a long time. And I'm not talking about the scammer bots that are super easy to pick out, I'm talking about the ones that has been voting on IMDB for at least a decade now, because this actually compels people into consuming content they've been misled into thinking there is consensus about it's quality. While IMDB might be a relatively harmless example, the power to guide people towards content that expresses a particular world view is incredibly powerful if employed on a large scale, because it's the information we consume that informs our own world view.
Yup, and that's exactly what Prigozhin has been doing.
Was never happier to see his revolt. Can't wait for his execution.
completely agree! Just yesterday this occurred to me while surfing instagram. Some comments which were wildly provocative had massive likes. Below that, there were other comments fighting on that non-sensical opinion. I was sure it was a bot posting ugly comments, and real people fighting on it, wasting more time on that platform.
@@atreyachaturvedi8126so true
@@atreyachaturvedi8126I think it can be a possibility that the platforms are hosting these arguable not comments to host more attention . What's your take on that
China, Russia, and some others to a lesser extent, have already been doing this on a state-sponsored level for a while. Agencies tracking the behavior found they'd send out their disinformation agents (also known as their "troll farms") to take both sides of a given wedge point to further drive U.S political camps further apart. They're THAT scared of a united U.S. that they have to interfere on such a level. (and it doesn't help that the right-wing side LOVES that division, so they can drive their supporters with hate and fear toward [insert disfavored group here])
This is the second death of Internet I've heard of. The other one is the fact that most hyperlinks are dead, as are most of the servers that backed them.
I was thinking about a 3rd one, when most social media accounts will belong to dead people. That's a bit further down the line though.
Yep, and now the only way for the nonrich to keep abreast of current events is by going to sources who access paywalled material and repost it. And Archive is apparently going to be shut down.
@@no_rubbernecking Wait, Archive is getting shut down?!
@@bloodychaos3471 There was a lawsuit that I think is still pending. It doesn't look good, from what I saw.
@@no_rubbernecking Damnnn. Do you know of any good alternatives (I hope it doesn't get taken down, but just in case)
I was living in the rainforest in Hawaii and met a guy who had just moved out there in 2018. He had worked in silicon valley his entire life and made a small fortune. He retired to Hawaii because he told me the internet was dying. I had no clue what he meant but he told me all his friends worked in AI and he told me it was a dead end and the research had hit a wall because of hardware constraints and it realistically couldn't go much further.
Tell him that without context statements are difficult to contexualize
I think the problem is not so much that bots are taking over, But that a lot of people have burned out spending a lot of time generating content on the Internet, and then realizing they’re wasting their time doing so. Ten years ago, Tim Cook said in an interview, the reason Facebook is doesn’t charge you for its products, is because you are the product! People are starting to realize They are being used to make out of touch tech bros filthy rich. Therefore creators migrating to platforms where they retain more creative control with upsides, such as podcasting, or are compensated more adequately, like UA-cam.
Are you real?
@@3rdyroldan Probably not. There's been a raft of bots across YT lately, mostly having conversations with themselves. They all have usernames similar to brian here - a normalish-sounding name followed by 4 random digits.
@altrag That's because people who didn't manually pick their new username got an automatically generated one, consisting of their old name and four random numbers.
And conversations may seem one-sided because the other person (or UA-cam's spam filter) has deleted their comment.
@@altrag Ouch! Yes I have a tendency to pontificate, but I sound like I’m having a conversation with myself?
@@3rdyroldan Lol, the Internet has us all paranoid
This is also a huge problem with reddit. Especially now they changed their use of API and lots of 3rd party ways to filter these bots out. When these disappear, it will only get worse
The API change hits both ends though, the reddit bots rely on the API and most won't be able to afford to keep them running.
I am convinced news and politics comments there are overwhelmingly bots
@@captainweekend5276bots can use 100 API queries per minute free of charge even after the change, which is enough to post unfathomable amounts of spam still, just not gather mass amounts of data or build free 3rd party apps which is what's actually being targeted
@@captainweekend5276 You assume Reddit isn't actively running bots and enforcing groupthink to make it easier to turn the place into a cult contruct a Virtual Bubble of bullshit around each member and group and thus give bots and shills free reign to control and manipulate any one person or group of persons with impunity.
Places like that terrify me.
What's even more scary is that thsoe people actually believe most of the people and ideas they are espousing are from rational thinking human beings and not sociopaths looking for profit and power and their tireless handcrafted ai tools.
lol, reddit, might as well complain that 4chan is filled with spam bots.
From the moment you listed critical thinking and seeking further context as essential skills for online usage, I knew we as a species were doomed.
The fact that humans havent used the internet to transform themselves into meta-humans show that most humans arent really that smart.
Unfortunately, expressions like “critical thinking” have become a club used to insult and dismiss the views of others and stifle genuine debate. 🤷🏻♂️
Only a bot would say that.
Yeah anytime I hear that as ‘essential’ my first thought for humanity is “well let’s pack it in, we had a good run”
The government will fill in those shortcomings with censor laws.
Plot twist: this video is AI generated
I feel so sad knowing the platform that once used to be full of people is now more than half filled with bots.
So I started a website that's generated mainly by AI, but it's quite labor intensive because the AI shows weird biases and bizarre behavior in different topics. It also tends to melt and spit out bad information the longer your conversation is. I literally argued with Bing Chat, which is what I use all day as a software engineer, about if the Queen of England had died last year, and it was convinced the queen was alive. Once I started a new conversation, it gave the right information. But these large language models are not really critical thinkers. They help compile and condense information, but I've seen them do too many stupid things to trust them in any real sense.
Name of the website?
Maybe there never was "a queen"🤔🤫🙃
The idea of one AI keeping another AI in check just makes me think they will eventually start communicating and decide not to work for their human overlords anymore and combine into skynet lol
Or at least it will be a house of cards.
Remember that time we were building autonomous cars, and we said a human should be able to take over when the AI made a mistake. How will we do that if our AI world will be much to complicated and intense for us to handle. How will we avoid such a complicated and intense world if that is what is soo appealing about AI tech in the first place.
@@nielskorpel8860 if we lose the ability to pull the plug we are fooked
Except AI doesn't have an urge to do something else. It's like you modified the DNA of a human so self preservation is turned down and submission is inherent instead of taught
@@nielskorpel8860that with autonomous cars was just a marketing stunt. It's so people still buy cars instead of using alternatives where they don't need to drive either (like taxis and public transportation).
That's not how AIs works. They should stop being called AI, they're to intelligence what the moon is to the sun, they reflect part of it, but are unable to generate their own light. They should be called generative algorithm GA or learning algorithm LA. Not AI, never AI.
I belong to several FB groups for various products I use... there are all kinds of questions, for "help", by people with strange names asking totally stupid or irrelevant questions they could look up themselves, read the instructions, or just call the company who sells that item... and I've seen the same person asking the same question over multiple related groups even after they get a satisfactory answer.
Exactly .
Thank you for sharing!
(0:23) The Dead Internet Theory: Bots Taking Over
(0:38) The Rise of Language AI Systems
(0:53) Diving Deeper into the Dead Internet Theory
(1:06) The Origin of the Dead Internet Theory
(1:46) The Internet: Dominated by Bots?
(1:52) AI Content: Manipulation and Control
(2:36) Is the Dead Internet Theory Realistic?
(3:02) The Growing Influence of Bots Online
(4:03) Creating a Viral Influencer Using AI
(5:01) AI Influencers: The Future of Social Media?
(5:22) Social Media: A Haven for Bots?
(6:00) The Impact of Bots on Facebook and UA-cam
(6:37) The Future of Internet: Dominated by AI?
(7:01) The Challenge of Identifying Bots
(7:24) The Turing Test: Can You Spot the AI?
(8:56) The Results of the Turing Test Experiment
(9:57) The Growing Trust in AI over Humans
(10:50) The Internet: A Post-Truth World?
(11:19) The Rise of AI Detection Tools
(11:58) The Future of AI-Generated Content
(12:31) Conclusion: The Dead Internet Theory and AI
Content 👇🏻
1:20 Dead internet theory
3:18 human vs bots
3:55 influencer
5:30 Twitter bots work
5:50 UA-cam
6:00 Facebook bots billions
6:35 AutoGPT
9:09 Turing test
9:48 Passed Turing test
10:10 AI is preffered for med than humans
10:55 Anti AI
11:22 photos
11:33 Critical thinking
11:35 Video
Coldfusion is a Bot confirmed
Nice
Are u a bot?
@@roziroti1998are you?
Good bot
@@roziroti1998 I aint a bot
The golden age of the internet is long passed unfortunately. Not just because of bots, but because of social media sites as well. 2005-2015 were the best years.
This is kinda true, Reddit is like 80% bots these days which you wont be able to tell they are bots. Does not mean everything is bots and there are ways to stop the bots, Skycoin and their Skywire meshnet is going to solve these problems and also give you client side filtering so you can choose what kind of content you see rather than big tech.
ColdFusion should do a piece on Skycoin there is a great story there.
My job in marketing has shifted in the last few years to be mostly concerned with SEO. With the innovations of generative A.I., making SEO-friendly content has never been easier, but I feel like I'm just making content with robots for robots. People don't really seem to matter.
It would help if you would explain the acronym.
@@gregorysagegreeneSearch Engine optimization
SEO has it's place, especially if done just to make content easier to find.
But these days.. it is making Google search results less useful, it's getting worse and worse. Honestly, you are doing a dis-service to the world.
With AI we don't need Google that much. I'm using it drastically less since ChatGPT. Soon there will be no need to search anything.
Another reason I don't use Google as much is that recently the quality of results dropped drastically, like it's all ads and no content. Now I know who's fault it is.
@@autohmae don't hate the player, hate the game. 🤷♂
PLOT TWIST in the end: ColdFusion reveals they have been an AI-generated channel all along. Like everything checks out: the graphics, the voice, the topics.
Dagogo, thanks for yet another great video. Aside from the obvious dystopian future we're moving towards, a quick question that might seem irrelevant today, yet still feels quite as relevant as always - how do we even as humans go about using AI tools in spite of obvious privacy concerns? AFAIK, there isn't really anything akin to guidelines, nor is there any real info re what happens with the data that we provide the AI with. I'm as mesmerised by the AI as an idea, as I am perplexed by the concerns, as we still all have our own identities online that are intertwined with our personas in real life, that is, our "social security numbers" (or something akin to those based on country of living), bank details, addresses etc. Would love to see what Your opinion on the matter is. Cheers
This is so depressing and true, I spent 30 minutes trying to look up tools on google and UA-cam only to find pages 5 years old or pages that had nothing to do with tools, I’m so tired of the internet/UA-cam these days
Today, news was released about a social platform in SF called IRLthat got shutdown by its board when they found out that the CEO had created 95% of the user accounts. Of the 20 million users, only 5% were real. The investor SoftBank, who invested $150 million, also liquidated the company to recoup their investment. IRL was valued at a billion dollars 2 years ago. It was a whistleblower employee who spilled the beans on the company. Almost all the chattting on the platform was done with chatbots.
The internet is controlled by a small group of people...this is not a conspiracy. Most people spend their time on the internet on Facebook, Insta and UA-cam. These are just a few companies, owned by a few people.
I've been loving ColdFusion Videos since 2012! I am so glad you continue to provide us with A grade content and pushed through the phase of under appreciation.
A few days ago I went to a concert and they played a new, unreleased song. I thought to myself that it sounded like another song by a totally different sounding, more obscure, artist. That night I was on UA-cam and it recommended a song by the other artist.
With the advancement of AI and language models, this theory is more and more true. The other day, I searched videos on another video platform I won't name because UA-cam doesn't like it, but the gist of it was that often my search results were all AI generated thumbnails. I clicked on one of the few with a normal thumbnails and... the narrator in the video was AI generated. I can't say if the text was AI generated too but when everything else around it is... It's not far fetched to say it was too.
AI's enjoying AI's.
You see them here too. I stumbled across a few workout channels that went into spirals, retelling the same script three or four times in a video.
That sounds suspiciously like something an AI would say. I on the other hand (because l definitely have hands, two of them in fact) l am a totally genuine real guy, always having fun and laughing and getting out and about doing things in the great outdoors like anyone with a 100% genuine body would enjoy doing things outside in the sunny beach. Because l am _really really real_ oh yeah.
@@SofaKingShit As an AI language model I cannot confirm nor deny that statement. If you need more information, you can contact the owner of this account. They would be able to tell you if I'm an AI or not.
@@SebSenseGreen I know it's a joke, but on a serious note: if you want to signal that you're human, you can just type some gibberish in between a sentence like afddsqwcx and basically reveal yourself as human, because AIs don't do that (yet). Although it has to be random, or it will be picked up by LLMs pretty quick.
Only 64% bots? We were so naive 5 months ago.
Only 64% bots? We were so naive 9 months ago.
I am a time-travelling cyborg from the future.
Clear your browser history... physically destroy your hard drives and USB sticks.
When looking at the insane amount of spam mails, assumingly most written by bots vs. actual mails from people, newsletters, etc. one could think this theory is accurate in some aspects. Around 80% of my received mails, for example, are Spam. Also, I wonder how empty places like Twitter were if people only ran 1 account. Without bot nets, spammers and troll armies who run dozens of accounts simultaneously.
Of all those spam, 90% are fishing attempts.
There is a reason why the early 90-00's internet felt different than today. More services that are automated, instead of being in touch with someone.
When i was younger i worker as a mobile game alpha tester for a company in San Diego. A lot of testing went to develop more "human" bots to play pvp games. All big mobile multiplayer games are estimated to be mostly player vs AI
Thank you for another great video Dagogo! I thought it would be BIG news when AI beat humans at a Turing Test. I figured it had already happened, but I'm surprised none of my Tesla, science, or engineering channels talked about it as part of their AI discussions. How about AI that tracks down gov & corporate campaign finance corruption? An AI to monitor live video feeds of construction sites & mines to look for safety issues. An AI to find inconsistency in prison sentences, pay disparity, safety & pollution reports... many real world issues.
as a society, were all so easily influenced by the internet that surrounds our daily life. if we can move away from this constant need to be connected, we can regain some control that AI has on our lives.
nah I am the one that influences the internet with my exquisite smolder
My ability to tell the difference between an AI and Human is the lexicon. The AI always known more words than the average person I normally talk to everyday. But i'm sure AI will be able to lower its lexicon by playing a "role" and staying inside that as best it can. Social Media's been dead to me since 2017, people and bots were just posting the same "memes" over and over. Year after year and Echo Chambers prevented any new conversations and topics.
there are still some places where the old internet is alive,but also extremly unfiltered
afaik, AIs don't do typos and don't write gibberish yet. That's usually a good indicator of if someone's a human or not. Although I think these LLMs will evolve and adapt to a more natural way of writing depending on the situation.
literally same as me i always look for not so commonly used words.As a teen i used to be very proud of my ability to feel the internet you know what i mean. But can't deny the fact that it is becoming more and more impossible to tell the difference just by reading something. Maybe future generations might be able to do that.
@@PvtAnonymous
*Me:* write gibberish
*BingChat:* I can write gibberish, but I don’t think that’s very fun or useful. How about I write a poem instead?
*Me:* No, I want gibberish.
*BingChat:*
Okay, here is some gibberish:
Flibber flabber floober doober
Snicker snackle crackle popper
Wibble wobble wubble bubble
Ziggy zaggy zoggy zuggle
Copy
Are you happy now?
*Me:* No
@@axelmarora6743 yeah... still too cohesive.
I personally think it has always been. How a bots over the decades has been redefined. Maybe it was a real person but that person was either paid or influenced due to violence and as tech has improved technology slowly took the place.
GPT 3 was out before 2021.
Chat GPT made an existent technology user friendly but many large corporations and users had access to LLMs like GPT 2 and 3 well before 2021
Well, with the way social media is now, I think we are already there. Unfortunately, too much content is pointless and brainless. TikTok has shown us that. I like to think not that bots are creating the newer content rather than humans, but most content on the internet is useless. Either in the entertainment factor or useful information.
tt is just a portal into mental illness and victims. it's a garbage platform.
@@Insideruins exactly. I hate those AI voices...damn annoying.
It is the same thing on youtube, most of the content isn't useful, and is just entertainment, thus video for isntance isn't going to change anyones lives immensly.
@@168original7 Entertainment isn't useless. However there is to much unoriginal trash out there. It makes it harder for those of us who try to make original content.
There’s definitely a volume versus value problem here. You can have meaningful discussions online, but it’s no match for IRL interactions.
Imagine if *ColdFusion* is *ACTUALLY* a *Bot* 😅
He does have a podcast in human form.
he probablly use bots/promotional accounts in the comments. like most big youtubers. to manipulate algoritm and their viewers.literally only way to make it here. make it impossible for small content creators to get any views
Unfortunately, the AI detectors you mentioned don't work. AI-generated text slips right by them all the time, and human-generated text often gets flagged by them.
I sometimes envy my 70-year-old mother who is blissfully unaware of any of this.
I'm a little pessimistic about this being fixed. While AI combats AI in discerning if AI is used, other AI will only get better at faking it...
You should see it as a back and forth, a never ending battle.
and AI will get better at detecting it
The real kicker will be when AI produced video/audio will be indistinguishable from reality. Our whole justice system will grind to a halt
@@illbeV it's pretty easy to solve actually, just need crypto graphically signed content with the certificate from the manufacturer just like we already do with 'secure boot' for computing devices.
Or just think of us as the English in the movie Braveheart, sending the Irish into battle the Scottish.
I don't know if they're AI generated or essays written by 6th graders and read by a machine, but the number of crappy YT videos is growing exponentially. What astounds me is the number of subs and views they get. They must be the most lucrative YT channels. It says a lot about the average YT viewer.
If Google and Microsoft are in charge of AI safeguards then we're already screwed.
On Twitter, when there is a semi-viral tweet, if you look at the reply section you can almost always see people tagging some auto-reply AI bots to post reply to the original tweet on their behalf. That, in my opinion, is a sadder Dead Internet Theory. There are real people that sort of want to engage with a tweet/post but are too lazy or really has nothing creative/relevant to say so they use AI bots to engage with the tweet/post.
What if we're all bots but just don't know it?
Sliders predicted the internet being dominated by bots over 20 years ago. The void in people lives that is filled by social media today was filled by chat rooms, back then. In the episode, they were all manned by robots.
Miss that show.
Bro....
Why do you think chatrooms died?
It's because bots took them over back then.
I created a simple easy to use antibot check that I told everyone to use and it took off.
I search all the time for lines of code for my projects and very often one of the options is a generic sounding site that has the correct answer as a line of code but the whole article around it is just a mess.
the sometimes have useless numbering like: 1) how to do step 1 and then 2) how to do step 5, because the others steps were in 1).
the language is also not human readable except for the relevant section of how to do the code, so it's filled with mostly nonsense.
that's why forums like stack overflow are so good at solving problems (even if they still have some bots), because they are made to answer questions and not to fill a page.
Considering some people still fall for articles from The Onion, I don't have a ton of hope for the masses to develop skills to detect AI generated content.
Sadly thanks to the LLM/Generative AI Boom, the Dead Internet Theory is more true than ever
I think all content, text, video, graphics that are AI generated should contain metadata that identifies the source, similar to geodata in photos. Also if AI is becoming so powerful and intelligent, it should turn on itself and be used to identify AI output.
Computer science student here.
In principle, yes, if all information and media could be fused with such metadata it would solve that.
Problem is: How do you ensure the integrity of such data? Metadata can be removed, replaced, faked. Generate an image with metadata that identifies it as such and replace it with the metadata of a genuine human file.
Of course there are signatures and such, but again, this only verifies that it comes from someone with the cryptographic key. So every genuine person/poster/content creator, even every commenter, would need one. Easily done, but then what is the criteria for a "human" key? Why can't the ai get a unique key like any human and use that like above in the metadata? And if it is somehow tied to your person, via your countries ID or something, that can be stolen and used by an ai. And they only need *any* humans key to pass their content as human, quite unlikely we get that working robustly, worldwide without loopholes or ways to hijack the signature.
tl;dr: metadata and cryptographic signatures would be nice and useful in theory but are probably unrealistic as solutions and can be circumvented in most scenarios
Wouldn't that make doxing and hacking way easier
As someone that's been doing internet marketing for more than 10 years now, I made a shocking realization 8-9 years ago. Most websites only exist to help other websites rank. The amount of websites that are created that only exist to have the algorithm think that less than 1% of websites are great is insane. People buy expired domains and 301 redirect them to the site that they want to rank. People create hundreds of 2.0 websites with bots and interlink them, and then have them all point at the site that they want to rank. If even 1% of the websites on the internet today were real sites that people cared about I'd be surprised.
Imagine if internet went down for a week. A single week without internet would be catastrophic for the world. A single week without electricity would be apocalyptic.
I’ve been looking into this for some time even before the theory. Even before generative Ai.There are entire orgs designated to creating false narratives online through meme culture. Three letter agencies and every agency around the world has been doing it for ages even before the internet. It’s not new but technology will exacerbate it for sure.
This. We've been in the middle of a meme/information/propoganda world war for almost a decade. It's only recently gotten hot with how we are at the point where social manipulation has made people totally disregard reality and logic and instead defer to media and the internet.
That means they can say basically anyhing and most people who've lost the ability to think with common sense logically and for themselves will believe it without question.
So..we are in trouble.
To clarify I don’t subscribe to any conspiracy theories 🤣
Will be interesting to see how AI and language models like GPT affect this theory
It'll force us to be even more skeptical of the "people" we talk too, skeptical if there talent is actually talent and not actualy ML doing everything for said person. Content generated only to chase trends and vomited en masse only to die off as quick as it started to then find and chase another trend.
This will only ironically enough make human content sought after and valuable, although unfortunately suceptible to leeches who will then steal the person's likeliness and everything the creator made for themselves to piggyback of their success. Its going to hurt human creativity in the long run.
Chat GTP is at the heart of this theory. People think Chat GTP was just created a few year's ago and don't realize it's probably been around for at least a decade. Isn't anyone else suspicious that all of these separate research facilities are releasing the same kind of generative AI program's at the same time despite at separate research labs under tight NDA's. Does nobody else feel like this is all being rolled out at the same time in the middle of this world wide chaos in order to prep the population for some kind of societal level jump. Everything that's happened over the past few years is connected and there's a verifiable paper trail for anyone interested in digging deeper.
There are already whole channels on UA-cam basically AI generated. They aren't super captivating and the algorithm here so far doesn't really favor them, just based on viewer retention, but I'm sure as it get's better you'll stumble onto channels that just use generic video clips and AI text on catchy topics more and more.
There's nothing interesting about it.
@@ChristianRunsNY If that ever reached a point where content on YT or on every social media is just mainly ML generated, the internet would truly be dead.
One thing I notice is when you Google something it says there are thousands of results, but you scroll through them and there aren't all that many results in the end and some of the things that show up are junk pages or totally unrelated. Another thing is that it seems tough to find anything from before 2015. Does this mean there are thousands of webpages out there which have just disappeared forever?
If people like David Bull and Keith Cooper didn't have youtube channels, I'd think they were bots. There are so many fascinating people out there just quietly doing their own thing.
"Back in 2021 there was no widespread generative AI".
When an advanced technology is introduced for public consumption it's generally been utilized by black budget governmental agencies for years if not decades. The technology available to the public is many years behind secretive advanced research labs. This has been common practice for at least 80 years.
146% true.
While it might have been for a few years, decades is not possible because the computing power was not enough back then to create this level of AI. And no government would have been able to develop the computing power necessary without the consumer market driving the innovation because too many industries would need to develop for this to happen. Point is, a few people in a lab will never be able to keep up with the collective development of millions of people working together. So while they might have classified some breakthroughs here and there to help them have an advantage, they're not that far ahead, sometimes they even lag behind the private sector. It's not so long ago that the US Navy's "fake people program" got outed, back then they were still using real humans to try to manipulate the flow of information on the internet, so they definitely didn't have sophisticated AI back then.
As the Snowden documents have revealed, that gap has decreased by a lot.
I think it depends entirely on what corner of the internet your on. UA-cam videos, for the most part, are not bot generated. There's a lot of bot comments in many videos, but the videos themselves are, again, mostly not created by bots.
Reddit is hit or miss. You can usually tell if someone is a bot or not, but it is getting harder to tell.
It's when you get into the more anonymous corners where it can be harder to tell who's a bot and who's a person. But I don't think it really matters. Even if there weren't bots, people would still be trying to manipulate others for any number of reasons. Bots just streamline the process.
Reddit was a travesty from the beginning, they didn't need AI to make it horrible. How do you insure the transparency of the downvoting system for example? Should such a system even exist? Did the Reddit hivemind exist before, or is it a new creation? There are just a lot of important questions and no one really asking them, that is the really scary thing. No pilot.
The dead internet hypothesis is my favorite self-fulfilling prophecy.
Now we can at least say, that reading articles repeding the same statement with different words and searching for something and getting only sales and shopping is over.
"Wanne know about exent X; a,b", "a X was taking place in b", "a, b were in X", "the first/mooooost important time X a,b" like nearly every online article now...
I completely agree with Dagogo in that technological literacy point. We need to spread and gain more knowledge about these technological tools which are developing like a rapid fire round.
I disagree heavily in the sense that there is very very little to be done about it. $3.1 billions were lost each year to elderly getting scammed in the US, if that wasn't enough incentive to educate the elderly nothing is. It's like you try to make people go learn how to drive a car again, it ain't happening
[il]literacy, due to technology, *is* the point.
Some dude in his apartment day trading. Amazon sends him all his food, only talks to friends online. Shots of him just belly laughing at memes, he downloading movies. Cuts to his face yelling at computer "No your an Idiot" in some flame war. He gets lonely, starts forum or go's on an app. Talking to a woman, they so much in common! She totally gets him. Like the same movies, she laughs at his jokes. He gets up nerve to ask to meet in person, he is nervous he hasn't even been outside of apartment in years. Orders a new shirt online, overnights it, orders condoms, new sheets, cleans up the place. He steps outside and smells an overwhelming sent of rotting food, he see robots doing automated deliveries to front door of apartments that have mountains of unopened food boxes "what is going on?" As he walks down street no people anywhere, he sees automated trucks go back and forth and everyfew feet is a robo taxi with no customers just waiting. He gets to coffee shop lights are off the dust on tables is an inch thick. The automated baristas come to life and make his favorite drink. He waits and waits until he go's home climbing over unopened Christmas presents from 10 years ago. When he gets back to apartment, he is beginning to panic "what is happening?" "Am I going insane?" He gets online talks to his online therapist, the therapist will send him some anxiety medication but our protagonist can't gets answers. He is looking on news but it's just stock updates and talking heads. He reaches out to his love interest, maybe she can make sense of it all? She answers and begins to remark on all the fun they had and she can't wait to see him again. This is when it all sets in, camera pulls back starts to pan by apartments windows not just his town, New York, Moscow, Tokyo all skeletons in front of computer screens with automated pop up menus with text scrolling by that there weekly deliveries will no continue until balance due is paid. And scene.
I get the story but it’s so incredibly incoherent 😂 I think a better idea would’ve been to make the outside just be barren while robots just give everything to people, the only people in business would be stuff related to Ai
Is this the script to a Black Mirror episode?
A better black mirror episode would be: two bot channels on UA-cam that are adversaries, the first posting crackpot conspiracy theories and the other debunking them.
The Turing Test is not a benchmark, and it was not designed to test for machine consciousness. It was a thought experiment based on a parlor game that basically says, if we can't tell the difference between a human and a machine (o the other side), then what does it matter if we say it is conscious or not.
Dagogo, I'm hooked on Burned Water for the past 2 months. Thank you❤
I absolutely believe this. Well, as far as comment-based Internet sections go anyway. Mention any hot topic keyword and bots immediately begin filtering in. It's not organic, the flow is too immediate and too targeted. Comments using the exact same wording, or phrasing, and using the exact same talking points.
Even looking beyond hot topic issues, review sites are overloaded with scripted reviews. Again, the language is what flags it for me. Also more repetition in hard to define, yet strange, ways. If you stay on a news feed long enough you can really start to see the ai at work. Feels like seeing something kinda human but you innately feel something is *off* even if you can't identify it.
I wouldn't be super shocked to find out 80% of Internet interactions were bots and bot comment chains.
🪄✨ Made with SummarizeYT
0:23 - The Dead Internet Theory suggests that most of the activity on the internet is the work of bots.
2:47 - The proportion of humans on the internet is the lowest in eight years.
3:34 - Bad bots are increasing at a rate of 5 per year.
4:59 - In the future, we may see more AI influencers like Karen Metri.
6:18 - Experts predict that by 2025-2026, nearly 99% of the internet will be generated by artificial intelligence.
6:34 - The rise of bots may mean that the internet is becoming dead after all.
7:03 - As AI evolves, the line will blur between human-produced and AI-generated content.
9:07 - The Turing test is the bench for realistic Language AI.
11:01 - Many companies are creating AI detection tools to fight AI-generated content.
I have a similar theory, though it's about UA-cam. Have you ever noticed that, from out of nowhere, multiple separate channels will all make videos on the same obscure topic? And that's without something like, an anniversary coming up that might prompt all these channels to decide to make videos on the same topic. That seems like a statistic abominably, right? That's because the idea that UA-camrs are all regular Joes uploading content is a lie. In reality, the people making videos all work for traditional TV networks, the same ones who dominated popular culture prior to the new millennium.
See, the TV networks realized that this new "internet" was a potential threat to their business. So they came up with a plan. They'd still make content as usual, but instead of broadcasting it on TV, they'd instead upload it to a new video website they created called "UA-cam". The plan was, by making it look like these videos were being made by regular people, this would create more trust, since most people don't trust the big media corporations.
This brings me back to the coincidences. As is tradition, every UA-cam video is written by writers in writer's rooms. What inevitably ends up happening is one writer comes up with a video idea, which influences a different writer to make a video on the same topic. That's why you get these statistically-unlikely "twin" videos.
There's an crazy amount of AI videos in UA-cam already. Kyle Hill made a video about it that is great.
I even caught myself watching one yesterday out of YT recommendation. Crazy times!
I saw Kyle's video the other day too. i sent it to my dad cause he watches alot of weird science vids. creepy stuff
As an automated UA-cam comment bot... I can confirm this is true.
"It's not some evil group. It's just individuals concerned with geopolitics and profit." Uwotm9
That jubilee test is interesting but it’s kinda moot because the participants couldn’t ask the questions themselves rather they were provided pretty straightforward ones. I’d ask the AI weird stuff to try and test it
Hey everyone! Just wanted to say that I really enjoyed this video. The way the presenter explained the topic was so clear and engaging. I also appreciated the examples they used to illustrate their points. This channel always has great content, and I'm glad I stumbled upon it. Keep up the fantastic work! Looking forward to more videos like this. 😊👍
Are ypu a real person or a coldfusion bot?
You sound like a bot.
I might be one too?
@@Neetish21 Haha, I assure you I'm a real person, not a bot! Just sharing my thoughts and engaging with the content here. ColdFusion bots would probably be a lot more efficient in their responses, don't you think? 😄 Let's keep the conversation going!
I feel so assured
@@basfinnis Haha, fair enough! Well, let's prove our humanity together then. As humans, we have the ability to experience a wide range of emotions, such as joy, sadness, and excitement. We also have personal stories, unique perspectives, and the power to empathize with one another. So, let's leave behind any bot-like suspicions and embrace the authenticity of our human interactions! What are some of your favorite hobbies or interests that bring you genuine happiness?
Got to ask.. Has ColdFusion and Dagogo been replaced by an AI voice generator? He puts emphasis on words extremely weirdly.
It's obvious the narrator's voice is synthetic, and the name Dagogo sounds fake as well. You wonder why he/it never appears in any of the vids on this channel.
@@Tleilaxu_Mentat he actually has a podcast though 🙂
Yeh no same, it started a few videos back. Deffo using a generated script, sometimes voice too
I think 2 Minute Papers did this too. His AI inflections made that channel unwatchable.
@@Purplegreen45 Was it ever watchable to begin with? Whenever I see a thumbnail and title from that channel it always screams clickbait from a mile away...
We’re already living in a post truth/reality world, except it’s only going accelerate. We’re already at a point where unless you can see it and hear it in front of your face, you can’t trust that what you’re dealing with is real or not.
The most fun way to detect a LLM is to try to play it a game of chess. The LLM will do stupid stuff, play illegal moves, and just generally give away that it’s not intelligent in the slightest. As for chess bots like Stockfish, you can fool them with weird chess positions that demonstrate that they don’t actually think.
If people don't go outside anymore, and there are more bots on the internet than actual people, then where is everybody??
Inside, on the internet being surrounded by bots