What scares me more than AI spamming everything is how many people apparantly take information at face value. Its always been necessary to vet whether or not something is true.
Honestly, a large portion of the human population has ALWAYS lacked critical thinking, this is not new, it is simply much more easily noticeable since everyone has a smartphone in their pocket. So anyone can Google something and just spout the first thing they see that aligns with their cognitive bias, word for word. Back before the internet, people were still doing this same thing, but were instead racing each other to libraries or centers of information to look up something that aligns with their cognitive bias. What many people lack is critical thinking and this can most assuredly be attributed to education. By creating a universal education system online that prioritizes and teaches critical thinking skills, more people can learn to have a healthy amount of skepticism.
we've been lied to all of our lives, since the beginning of organised society So, even if it wasn't true, at least it was human. At least you were being lied to by a person I think the fact that you might be talking to a machine is what ruins people's trust
There's a poetic justice to the idea that the reason the universe is silent is because the aliens are all afraid that if they speak up a galactic digital overmind is going to start spamming them with messages about how you can get a bigger glubark with this one weird trick.
Ooh, I just got an interesting story idea! There's a rogue super AI whose sole purpose is to learn and communicate with another species. However, it's hyper efficient to the point that it floods all communication systems in an attempt to mimic and verify the model's accuracy (not considering time perception differences). This renders a civilization's technology useless because no piece of tech can actually understand what's being said. Ignoring stuff like faraday cages (or not?), it would be interesting to see how a society develops workarounds. Or maybe some stupid alien species did something very similar AND programmed the AI to replicate itself onto that civ's systems. Its sole goal is then to farm the species' creativity while limiting technological growth.
@@kellymoore5517I think memes prove the opposite though. They’re anti-intelligence. Like ppl creating infinite car bumper stickers with a catchy phrase.
@@shitmandoodNot necessarily anti-intelligence in the slightest, just shorthands and simplifications of concepts. I mean, we have this pretty sick meme called "science" going on...
What’s alarming to me as well is the people who should be writing legislation to regulate this (or attempt to, anyways) have no idea how the internet actually works, what AI is, or what AI is capable of.
Thats why there are supposed to be experts advising them. Theres a whole army behind the scenes of analysts/ subject matter experts/ etc. but the policy process is definitely slow
To be fair... its because people ARE getting dumber. We've been increasing the mount of "outsourcing thinking itself" as tech has advanced... all of this BEFORE LLMs popped off. If only it wasn't by design so that humanity can be dumbed down to a level that makes them easy to enslave and control.... Oh well.
For me the worse "articles" can't even call them there is just ones that report what someone said but gives no context of what, when or if what is being said is true or not, I wish news articles should at least give more context than just parrot what is being said to them without a thought, worse when they are quoting someone clearly lying even with video evidence in the same article showing what they are saying is completely false. In my country a book is being banned from middle schools because it has a sex scene, which I think is weird because many middle school books here have explicit sex scenes and one I read as a teen also had an explicit rape scene, and these books never received any "Think of the children!!" arguments to be removed before, to make things worse the politicians that are banning said they never read the book they just were informated that the book had pornographic content and wanted to ban it, and what the news articles are doing? Just quoting these politicians that never read and now many people believe is full porno and I always when people start talking about it I ask what passage they think was too strong for middle schoolers and they never know, they just say that most be true that it should be banned because many people are saying its bad... And this treatment go to anything is even for health, like many news sites spread fake news because they just quote a person and blast it in their front pages, like had people stopping their cancer treatment because they decided to quote a priest saying that the best remedy for cancer was tea.
If I would have known it would become this current dystopian hellhole, I would have never taken the old internet for granted and enjoyed every second of it.
@@dividedstatesofamerica2520Same. But I do cherish those good memories. It's really bittersweet, but it makes me feel good when I think of the mid 2000's
As a security researcher, Dark Forest of the Internet is the best way I've heard so far to describe the kinds of ubiquitous threats the average user faces online. I'm definitely going to borrow this for educational purposes and credit this video!
Right? The 'Dark Forest' analogy really hits the nail on the head, doesn't it? It perfectly captures that vibe of the unknown lurking around every corner on the internet. Using it for educational purposes sounds like a solid plan. It's a stark, visual way to highlight the importance of cybersecurity to people who might not realize the extent of the threats out there. Maybe it could help demystify some of the more complex aspects of what we're up against online. I'm curious, though, how do you plan on incorporating it? There's so much potential to expand on that analogy, especially with how rapidly the threat landscape evolves. -This reply written by ChatGPT4
@@elliotgillumPerhaps, but Kyle’s video is far more accessible and comprehensive, and Kyle also clearly states the theory is by Yancey Strickler. He’s fine. Also did you know Strickler is the CEO of Kickstarter?! I didn’t!
@@titiwa632As funny as this joke is, the purpose of René Descartes’s “I think, therefore I am,” is for someone to prove their own existence for themselves because the fact that they are thinking at all means they must exist. It is part of a philosophical argument to determine if everything we believe exists actually exists, or is just some illusion, but “I think, therefore I am” doesn’t prove someone’s own existence to anyone other than themselves, so it can’t really be used as a pass to prove someone is human, AI, or even exists at all to another individual. This message is AI generated. 🙂
There is something we did on our discord server. During onboarding we have mandatory multiple choice option before anybody can use the server. The first option says "im a bot, please kick me as soon as I join (dont choose this role if you actually want to join the server)" and it actually weeds out a lot of bots joining.
Same. Articles/studies I read 5-10 years ago and used for reference in discussion/debates/informing etc aren't showing up anymore. I don't think they've been erased, rather buried under curated info by corpos with agendas and adding AI into the mix is going to muddy truth tenfold. Kinda glad when this shit really hits, I'll be on my way out of the eventual dystopia.
thats why me and a group of people are trying to gather as much information as possible and put into our "archives" a AI free community where you can find all the original information. This isn't an ad or something like that, it is a project of the future, or better said...for the future
“Once, men gave their thinking over to machines in the hopes that it would set them free. But it only allowed other men with machines to enslave them.”
Sh*tposting is life. It's about the only entertainment we have left now our governments have basically turned us into the culture of demolition man lol
I'm beginning to understand why when my grandparents were dying, they didn't really seem to mind. The world they grew up in and loved was already gone.
Absolutely spot on. 🙋🏻♀️🇨🇦👍🥲At 62, I am at glancing in the distance at that exit sign thinking just that. Where did my world go? What world am I in even? Sometimes feeling invisible even feels comforting.🤷🏻♀️🐱
@geoffreylaxton671 Sorry to say, that I completely understand. Somewhere around 2012, it seems like I entered into some horrible, alternate universe, and I can't go back home.
How do we know you're not just an AI competing for top comment? How do you know that I'm not an AI? I will bibbly bopino the wambly trunator, that's how.
I must be getting old because "back in my day" we never took anything on the internet seriously. Rule 1 was "dont believe everything you see on the internet" and rule 2 was "never tell someone your name or address on the internet". I dont know why these rules were ever forgotten in the first place.
Funny. "...must be getting old..." In my experience it was my parents generation who tended to believe and forward any scary email they read. I stopped counting how many of my replies to their emails contained links to a Snopes debunking article
@@bornachexcept your parents aren't who he's referring to. Before the iPhone, before minecraft, before even internet gaming, there was a generation of users small enough to have a meaningful society.
They weren't. People were gullible back then and so far as I can just as many people remain just as gullible who fall for that shit vs people who don't fall for it. And no one gives out their name or address more than previously either not to the point its noticeable. People were gullible and people will remain gullible. Its definitely you getting old and looking at the internet with rose tinted glasses nothing has really changed on those two fronts at all. But also its weird to say no one took anything on the internet seriously but also took it dead serious enough to never share personal information. Sounds like taking it seriously to me lol
My grandma was nearly scammed with a synthesized voice of her grandson (my cousin) asking for money because he was in a car accident. Luckily she actually called him on his personal phone to make sure and she wasn't scammed.
@@JoshuaBenitezNewOrleans they need just a little bit of your voice. could be facebook. youtube or they called someone pretending to be some company and recorded the voice that way
I'm so sorry your grandma had to deal with that and I'm glad she didn't fall for it. We've decided on a family passphrase- something not recorded/written anywhere - that no-one outside of us would know which will hopefully avoid this situation. I think we're all going to have to do that with our older loved ones now :(
I remember an anecdote I was told during an information systems class “that if you are lost in a dark forest, the only way out may be through burning your way to safety.”
@@pkmntrainerred4247 I think it would be more along the lines of not using any sites that you aren't sure if the content and people are real or not. Keep going; deleting accounts, until you're back to where you're sure these are real people, thoughts, and ideas, or the outside of the dark forest. I think the dark forest is an individual psychological experience, so burning it wouldn't be something that affects the actual internet itself, but how you interact with it
Back in the early teens I got a phone call from some jerk. First, the call was very statically, but, and more importantly, this jerk claimed he was my grandson. I told him he was not. He insisted. I again told him he was not my grandson. "Yes I am!" he said, claiming to be in jail in California. I told him he wasn't meowing so I know he wasn't my grandson. Because I never had children, I had, and have, cats. This was not unlike the time I got a call claiming he was from the IRS, that I owed money and if I didn't him RIGHT NOW the police were going to be at my door. Well, first I had to pull over because I was in my car, and second, I knew I didn't owe any money to the IRS. First I challenged him as to what my address was. Needless to say, he could not tell me. But he kept pushing, so I told him he'd have to talk to my accountant, who also happened to be my brother. And I hung up because I wanted to go home more than I wanted to give him a hard time. I'm 70 and I just don't understand how and why people fall for this crap.
I think they’re trying to target people in the early stages of dementia :( The con strategy is unconvincing by design-scammers want to weed out those who can catch the scam early on. Those who fall for the first scam are then targeted for more scamming. So if you haven’t fallen for their scams, then you’ve been weeded out of their potential victim pool for now.
That's the whole point. They purposefully make their scams stupid so that only the truly uninformed, oblivious people will fall for it. Because these are the people least likely to fight them or be able to track them down and get their money back.
I got the "We'll have to send the police to your address unless..." call, and I said, "Go ahead." More shouting, cursing, the Indian accent of "Agent Johnson" got thicker and thicker...
I remember like 15 years ago, Tom Scott predicted that Privacy would become an outdated virtue, and Transparency would take its place. I think that was very prescient of him.
@@unsubme2157 forward seeing, or predictive. Also could mean foresight as in clairvoyant, but I think in context it just means his prediction was correct.
@@krishanSharma.69.69f The only way to define what something means is by knowing what it doesn't mean. The opposite also holds. By explaining the meaning, you also know what it doesn't mean, as they are pairwise disjoint.
Minor correction: Text based captcha are already mostly not used because they are no longer reliable. This happened a few years back and is the reason most captcha switched to just a check box or newer ones which are entirely behind the scenes.
@@BlazinLow305 Its more about just mouse movement. Its previous browser activity, cookies data, and all sorts of other information. This is why if you're in incognito or something, where cookies arent stored, you're far more likely to have to actually do more captchas after clicking on the checkbox. Its not so much that AI "cant" copy these things, its more so that these sorts of things take a lot more time and effort to copy well, making it more expensive to bypass. Plenty of bots are capable of bypassing these captchas, but the goal is to make it as time consuming and expensive as possible to make it not as worth it for bot farms. Most bots used in bot farms are extremely simple to save costs. Scams are trying to make a profit afterall, so higher operating costs makes them less profitable.
came here to say the same thing. blurred word captchas haven't been used in YEARS. that was kinda weird to see him so out-of-date with that bit of info.
I appreciate how you use the same straightforward tone and quiet music that you use (presumably out of respect) when discussing people dying of nuclear accidents. At the same time, it's quite chilling.
Eh another doomsday prophet. Only talking about the downside. Or ye, he added 15 seconds of "It can also be good". Do 15 minute video of the literal utopia that is coming and I will take you seriously. If not, you are just a doomsday prophet.
Our parents' warnings of "don't believe everything you see on the internet" have an eerie prescience in the age of algorithmically-generated content...
It's soon going to change to "Don't believe ANYTHING you see on the internet" but we all know there'll be people who will continue to believe just about anything they read.
Ironically it's often parents i.e. older people who're much more vulnerable to online disinformation. Cos they're not in the habit of being skeptical or checking stuff up. Their security came from trusting people they know. So all it takes is the message to come from a trusted source who's been hacked and they'll accept it without question. A lot of sketchy alternative medicine ends up marketed this way for instance. Despite it being called a 'curated community' here, Whatsapp is full of viral chain mail crap.
@@lorpen4535 This is basically what I said here, but YT in its algorithmic wisdom seems to have deleted my comment. But it's still sending me notifications of other comments. Trillion dollar company btw.
Synthetic or not, the corporatism and cleaner internet was what truly began the beginning of the end. From the moment someone tried to curate the internet to make it more marketable was already on the path of death.
I mean these problems are just as bad on search engines as social media. If search drove as many views and people spent the majority of their time on independent websites the bots would just make those instead. Making a website is just as easy as writing text to an AI.
@@rjohnm666 The beauty of that is that it's not always people. There's entire botnets dedicated to creating political discourse now. The arguments you hear may just be trained autofill bots, bots whom the sites on which they post will not counteract because they create engagement. It's insidious.
There have been numerous academic papers recently that have shown that bots are more effective at solving CAPTCHAs than humans are. At least for many of the common types of CAPTCHAs that are currently in use.
youre right and actually the captchas doesnt detect a well made performance in being completed but the opposite of that to recognize the human common behavior that is more erratical
There is a Dutch add that warns against posting your kids on social media and the dangers of AI with it. You should be able to find it on yt, show them that short add and then they will never ask you again
Even 20 years ago that was the right approach. Good thing you've been going that way. I worked at a free-hosting company around 2000-2001, and we regularly provided evidence of CSE rings that got them busted, but there were also things like a family innocently posting beach day with their kids, and some creeps found it, and suddenly their kids' photos specifically got 20,000 hits overnight. It's gotta be horrific for them realizing their kids' pics are being shared by pedos for arousal purposes.
I will never post my kids online if I ever have them. And from now on I will always put on my pictures a text bar "I do not consent to this image being used to train any AI model." If enough people start doing this, it will force the creators of AI models to stop scraping images from the internet, and force them to ask permission to use an image on their training data. Because it doesn't matter if a website says yes if its users (aka the people creating the content) are never even asked.
“Why are you hiding them from us.” Come visit tf? You’re always welcome.. ?! Besides there are always private family photos albums that you can post to… I can’t imagine being entitled to someone else’s only appearance.
The internet used to be the one place you could go when you were lonely - regardless of how awful you felt - and at least be connected to other humans in some form. Feeling that disappear is scary.
Maybe it's my age, but I think the restoration of a balance between 21st and 20th century interpersonal dynamics would be really nice. I don't think we'll totally lose the internet but I would LOVE to see the destruction of current social media culture, which I never fully embraced. All the oversharing, privacy violations, toxic trends, and shit like that, imagine the distrust being enough that one day, people just turn off TikTok. (An app I refused to get, BTW.) Some internet will still be good and useful but being pushed back outside would be a massive boon to our mental health on a societal basis.
Ironically enough, an AI therapist is or could probably be just as good if not better at making you feel less lonely. Some people already find AI therapists to be helpful.
A buddy and I were talking. He likened it to how phone calls are a dark forest already. It’s potentially hostile to answer a call from a number you don’t know. less scammers find out you're a real human. You only call text/people in your phone book. I think the surface internet is rapidly approaching a similar dynamic. This has precedence.
I never thought about it that way but that's actually true :0 I used to make a point of answering all calls because I'd rather get some spam calls than miss a single real call. But the few turned to many, the many to excessive amounts, the excessive amounts to unmanageable amounts of, and at this point not only do I trust Android's spam number detector, I don't even pick up calls from private numbers because in the last year, no exaggeration 100% of those has been spam calls. It's impossible and I've stopped trying.
I mostly don’t answer my phone if I don’t recognise the number or have it saved. It’s frustrating to near constantly get robo calls or calls from someone who’s clearly trying to scam you. Especially as a night shift worker, they wake me up with their nonsense 😒
As someone who's passed through everything from "You're the man now, dawg," to "Do you know de wey?" and on to gyatt rizzler, I gotta say I'm not convinced me or a lot of people in my age group could have any chance of catching up on the latest memes faster than data-scraping LLMs...
The latest brainrot memes could be the only way of human to human communication in the future, like a secret password to enter a human hideout, or a code, encryption of sorts to confirm you are talking to a human. memes change weekly and spread like wildfire. it can reach humans before it reaches ai.
AI can't shitpost the same way we do, especially with images because they cant look at it, decipher its meanings or what makes it funny and repost it sith actual feelings. So personally i doubt the bots can keep up with US when we have accounts entirely dedicated to making and reposting goofy ahh memes.
13:58 “Maybe Rizz and Gyatt will end up saving something uniquely human online” is not something I expected to ever hear in my lifetime but here we are
Humanity is defined by starting to massively use something new that it does not still fully understand and dealing with the concequences later. Lead, phosporus, radiation, asbestos.. those are just the ones i can come right now from the top of my head
Add fossil fuels to that list. And even though we know for a fact that the continued use of fossil fuels is going to end us, we continue with business as usual, because switching to a more sustainable way of living isn't profitable right now.
"No LLM will be able to keep up the pace with weird internet lingo and memes". This is a funny statement, I wonder how much time will it take to age like milk.
A certain and easy answer is when AGI is born. If one ask for my non-trivial prediction, then I'll say that it'll be so when photonic computing are used extensively.
We already have RAG models, they may not be perfect now. But it has the potential to prove that statement wrong. NVIDIA has already implemented into their own chat interface. Huggingfaces' chat page also has RAG capabilities.
An LLM with a long context window like Claude 3 could probably already keep up with Internet culture if it is given Internet access and runs in a loop.
I’m on Elon musks side with this. Someone was going to create this no matter what. If it’s not good actors with good intentions then it’ll be the bad actors who develop it. It’s important to explore these things regardless of the danger because of good people find solutions to keep it from being a danger that’s a whole lot safer than bad people having access to the technology and doing anything they want with it.
I think a lot of people fail to realise that AI doesn't have to be "smart" for it do be dangerous. Much like early computer viruses, a piece of code with a simple objective in an ecosystem not ready for it, much like introducing a predator, can be catastrophic. Add to this the genuine loss of tech literacy with generations who are growing up with tech "just working" and the ability to even understand how it's a threat at all diminishes. One of the most common issues I've had to assist students in troubleshooting, at a university level, is running every program under the sun all at one and a million chrome tabs, or not understanding that their device has a physical limit on storage. this results in complaints that our software is buggy or doesn't work for them, when in reality they fundamentally misunderstand the machines they use for most of their waking lives.
I think people also fail to consider non General Intelligence as an arm of human interaction. The machine may not have wants but it's operators will. It will be the mask they put on and the megaphone they use to drown out dissenting ideas.
You could try giving them the task of protecting the Internet by giving them a box with a red light on it (Tell them THAT is the internet and they must protect it!) Ref: The IT Crowd Season 3 episode 4
In middle school (not many years ago), peers would ask me for the time, so I would point to the nearest clock, only for them to ask again or ask to look at my phone for the time. This level of tech illiteracy does not just extend to clicking a few buttons on a device, but also to reading analog clocks, something we were all taught in kindergarten.
I’ve been a programmer for over 17 years, one of my biggest fears is when I inadvertently fixed a bug in my code and don’t know why it’s working. AI can generate code for you now, and this is a nightmare. It’s difficult enough inheriting a code base you didn’t write yourself, imagine inheriting one that NOBODY wrote.
It is truly horrific. I hope the people that create and support such abominations are caught and rightly punished. Personally, I think the only way to combat intelligent bots is with more intelligent bots created for the greater good, so hopefully law enforcement could utilize A.I. for the better. I'm not optimistic, though. To anyone who read this, stay safe and try to keep your private life off of the internet 🫂
"When's the last time you believed any headline, any social media post not wanting to check if it's truth / has the right context" Probably some time before I actually got a little media literacy, clickbait in media existed long before even youtube and it made blind trust in headlines a bad idea way before AI was a thing.
This completely. Propaganda and agendas are not something that was magically conjured into existence because of the internet or wasn't widespread because of it. It just made it easier to spread new versions of it.
Yes. AI lets us mass-manufacture BS, instead of relying on the handcrafted, artisinal BS that we previously enjoyed. The BS itself is nothing new; just the overwhelming quantity of cheaply made knockoff BS is.
Was gonna say, the 2016 election, and coverage of an unrelated individual, is what really woke me up to the blatant lying and deliberate misconstruing of context the legacy media is guilty of. Does not matter the ideology driving the specific organization, they're all guilty of it. At the very least, that has prepared me for what LLMs will do.
@@maxsalmon4980 All the corporate news just copy each other as is. Quantity doesn't matter when traffic is dictated by what big tech wants to spam you with.
The point still stands though. AI allows such a large amount of false information to be created so quickly and spread instantly. It's going to get to a point where you can't even check the truth or context of the original post, because every new thing you read while trying to search for the base facts will also be shrouded in fake posts and articles. And those fake articles will be so similar to real articles that it will be next to impossible to know which is which, unless you watched whatever event you're researching firsthand and don't need to look anything up in the first place. It's more than just being smart enough to not trust everything you hear, it's going to become almost impossible to even research things online at all. Our legislature, ways of detecting AI, understanding of AI, plans for how to handle AI, etc. are all lagging behind the actual proliferation and creation of new AI technologies and are only going to fall further and further behind if we don't stop to address it right now. As a Software Engineer who has taken multiple classes on AI and AI ethics, AI is a terrifying thing in its current state, and not enough people are planning for how to handle its progression.
memes also taught me how to talk about capitalism, transphobia, mental health (namely, mine), and the ongoing epistemicide of indigenous people. (fresh memes also make an excellent nazicide!) though the whole idea of there being effective methods of "combatting AI" (the possibility of boticide) is widely regarded as a pseudoscience among the bot scholar community, and generally sneered at in botiversity campuses.
As someone who has been using the internet since the early 90's and has made an entire career developing websites, I'm ready for the internet to just go away.
I'm ready for the internet to only be allowed to be used as an archive for libraries. I mean, the revolutionary thing about the internet was the easy access to information, so let's make it just for that. Only libraries can put information in, and we can only use it to search and access the information provided.
@@NankitaBR Putting information in... where? If your client computer can't upload data, you can't physically request the data you want to download and the internet becomes entirely useless. Most internet resources already are the way you suppose - curated content that can only be consumed, not added to except by whoever holds the keys. So what you're suggesting is basically turning off comments on sites that allow user input and engagement. At the least there's a deep irony in posting a comment to a youtube video suggesting this!
Great video! Just a note: distorted text and image recognition captchas have been automated with high reliability for a few years now. That's why there's new techniques such as align the puzzle, find the identical figure, etc. But even those are getting automated sort of easily nowadays. It's kinda mouse/cat situation.
Exactly there is incentive to make captchas only solvable by humans and in the other side there is incentive to automate the solving, it's an arms race.
from what I understand about those tests, they're really just tracking human imperfection. and while that can be emulated with little effort, if any solution gets used for mass automation, the method will end up tripping alarm bells and become unusable.
Have to be that guy, but those captchas are only being used to further train AIs. Captchas filter bots from people using mouse movement and cookie data.
biggest issue is that google is using reCaptcha as a training data source so anything they come up with to prove you're human will just be used as training data to make a DLM that is capable of solving it google in their greed is just shooting all of us in our feet
this whole concept reminds me CONSTANTLY of the Cyberpunk (ttrpg) concept that the open internet is so garbage and virus laden that there is a whole career based around finding things for the layman I feel like we're not as far from the need for Agents as I would like
Since Twitter became mainstream I was surprised the job did not exist, and since then more and more art critics and human curators were simply replaced by algorithmic recommendation systems.
R.A.B.I.D.S. is basically already among us destroying everything useful. Soon the only place with any valuable data left will be private networks...maybe we can call them data fortresses?
have only plaed the game (2077) and everytime i hear one of these news about AI all i can think of is about Bartmoss, the BlackWall and Netwatch agents Like, why is it happening, why can't it just stay in a game as a story, not as a warning?
@@Dragonflyer74 he had valid concerns but the way he acted on them was fuct. Even if you thought killing random and basically blameless civilians to make a point was cool, it was obviously ineffective, which makes it stupid.
7:46 there's something deeply existentially terrifying about the fact they requested a list of human traits AI can replace and it gave a response, and one of them being "empathy"
Fun fact about captcha. The test for it starts before you reach the captcha. It uses your internet history and mouse movements to determine if you’re a bot. Bots go straight to the website with no history, move very quickly and in straight lines. So in order for a bot to pass it they need to do unnecessary movements and purposely be slow.
If you train a bot on mouse movements from a data set of humans and give it a fake history of one of said humans it could pretty easily pass the click test.
@@asmosisyup2557 eh, not that easy. The variables and random directions are always mathematically perfect. To truly get the simulated randomness, you would honestly be better off recording a human's mouse movements and using those.
This is such a distressing concept for me. I already feel disconnected from everything and everyone in real life and being surrounded by an internet of computer generated content is going to drive me insane.
Take up a hobby that doesn't involve being online, like playing a musical instrument, knitting, model-building, etc... These activities can be quite stimulating to the mind without the constant barrage of bots.
As a generative AI model, I think everything will turn out fine. I'm just kidding, I hope you're not in too much distress over this stuff. I think the comment above is pretty important, thought. Taking up an offline hobby, specially one that makes you gather with people who share the interest, like chess or a trading card game, might do wonders at a time like this.
Maybe it will be a good thing. I see this to have the potential to ruin the internet as a place to just exist in as a surrogate for real human interaction. Maybe we'll all go back outside again
For decades I’ve used the internet to fill the social gaps I’ve experienced in life. I even met my wife online. But this video brings to the forefront of my mind something I’ve been feeling for a while but haven’t been able to place. The lack of humanity present on the modern internet. It’s a strange thought, that I may be of the only generation that actually found humanity online. Will my children understand that some of my strongest moments of bonding with others occurred in a forum? We’re entering the internet 2.0, and I’m not sure I’m ready for it.
he's talking to us from the future, when the last remaining humans live out there in the woods while the cities are populated by different AI interacting with each other
The internet has been dead for a while. With all the "SEO Pollution" as I like to call it, Google and etc services have been burying many things people are actually looking for. Not only that but Google also nowadays makes it really hard or impossible to find really old images and etc. See "The Dead Internet Theory" for more on this subject and then consider when this event took or will take place...
lol, do you mean glamping or is it just a glamorous way of looking at how your reality will become? I mean, not camping of course, but living in the woods, since we're now considered trash and don't deserve living in a city.
whats horrifying is that curated spaces will not be capable of filtering bots in a few years. Imagine advertising bots that spend years talking to you to gain your trust, capable of joining and talking in VCs and playing games and sending consistent essentially photorealistic selfies.
just to add to this, we already have rudimentary systems that can link multiple models together to do what chatgpt, dalle, and other models did together. Conceptually this already exists, its just not yet good enough to fool most people.
to be fair its mostly capitalism. if we had no reason to be afraid of corporations trying to use humans as capital, we basically would have no reason to be afraid of AI. if anything we would be happy with automation replacing labor so people would be free to pursue things that give their lives meaning.
"Disillusioned words like bullets bark As human gods aim for their mark Make everything from toy guns that sparkTo flesh-colored Christs that glow in the dark It’s easy to see without looking too far That not much is really sacred" Bob Dylan 1965 - This verse hits home for today's time.
Kinda hit home with how i won't respond to random messages unless i actually know who is messaging me. One thing I've noticed about AI is that it isn't very good at saying it doesn't know an answer to something. It always tries to answer.
Outstanding analysis. The real upbeat ending here is that we can all put down our phones. When I play terraforming mars with my friends and go to ace hardware and then work in my yard to build raised beds for my bulbs, it's all people... and there is no doubt.
Let's practice algorithmic incoherence... Regarding TMars. Blue's early, build engine. Prelude? So many fiddly bits! 3D printed boards for xmas: so much WIN! Got Turmoil. Hella complexity spike.
It certainly feels strange to live right now. We have just witnessed the explosion of something as impactful as the atomic bomb in just a few months, and we still don't know how to react
I used the same analogy when I wrote a paper about this 10 years ago in college. Not to sound self-important, but these past ~13 years for me have been like being tied to train tracks and watching the engine thunder towards us. There's nothing I could ever do about it, but I had to watch it coming. This is the first blare of that awful horn, and any who weren't looking before, surely know it's coming now.
I wonder if AI will cause our language to evolve into something similar to that world from TNG: Darmok. Everything we say communicated in referencial memes.
What do you mean "once" the internet is already infested with AI just do a google search and the first site will be ai generated and full to brim with ads.
Perhaps in regards to socialization and the exchange of ideas, but with every corp business practically forcing you these days to use their website or app, that opens the door to any number of issues. And never underestimate the amount of ridiculous treatment people will accept towards themselves (or others) in the name of status quo comfort.
I feel like that this will force two potential of many outcomes to occur: 1. The internet gets scrubbed at some point of the influences of A.I. somehow 2. A new "socialnet" or something gets created separately from the internet where only verified humans are allowed to interact with it, and it is ruthlessly guarded against A.I. content regardless of its origin.
You can't prevent the use of AI in your new socialnet... How can you prevent someone from generating text and just coying and pasting it in a page on your socialnet. The genie will not go back into the bottle.
The irony of the internet being the place that would bring us together, devolving into an isolating cacophony of meaningless content, is quite something.
This morning was an eye opener. First a post from an artists complaining about shite AI generated art instruction books, then my wife telling me about AI generated crochet patterns that can't be made in real life, and now Science Thor warning of the dark forest. Gah! Even the algorithm is involved! I'm going outside with my dog.
@@mateusvin well if i ask what did Kyle say in a specific part of the video and you answer that could work right? like what did Kyle say/do/show at 12:00? bots would have a hard time with that
@@shalimarsnow9316 I think it was Morpheus that asked, “what is real?” Electrical signals processed by your brain. I mean if we are all AI it doesn’t make our existence any less important. Real is whatever we make real. We still can understand when something doesn’t feel “right” like hurting another person or telling a lie. I don’t know. There’s still something to us, whatever we are.
Kyle I just want to give you a huge compliment and virual hand shake as I absolutely love your content. Utterly engaging, fascinating and educational. This alone has opened my eyes and honestly I found it difficult to comprehend some aspects of what AI is becoming, but thank goodness you explained it the way you did! Many thanks to you from Melbourne Australia :) - and yes I am a real human btw !
Captcha doesn't just rely on reading and reproducing distorted images of text. Some captchas are just clicking a checkmark. They actually track your mouse movements, browsing history, etc. to determine if you are a bot.
we also have these things and are very much trying to implement them to allow for data verification. Provenance isn't exactly a big deal on the bigger scale with c2pa and friends - but all that is sort of moot when you consider the current climate of nonsense: lobbying and anti-journalistic sentiment is already so high, all gen-AI is going to do is level the playing field, at least somewhat. Plus: who says I want to explicitly point out my humanness online - or talk to other humans for that matter? I mean, 99 % of people who have used an LLM to bootstrap their learning endeavors can perfectly put into words why they prefer machines over the real deal. Even great teachers can be real pricks at times, dealing with an individualized chatbot tailored to your specific needs, one that won't make you feel ashamed for asking easy or silly questions... that alone is going to completely revolutionize very "RL"-type of tasks. The Internet already was just platforms allowing you to curate specific feeds, reddit is literally called the frontpage of the Internet, and face it, it's not like the big subs could get any worse with synthetic content. Which you can still moderate, catching vague shilling or anti- sentiments isn't actually rocket surgery at this point. If all fails, new communities will crop up, but honestly, chances are we'll just enjoy talking to synthetic communities more. It already is way better than exposing yourself to Stack Overflow, among the most judgmental communities next to German pinboard culture.
When the discussion came up for how to confuse AI with cultural lingo and memes, it made me think of Kevin's "small talk" from the office. 'Why waste time say lot words, when few words do trick?'
Reminded me of the ST:TNG episode "Darmok" in which Picard and crew encounters an alien race who communicated entirely via meme which rendered their universal translator technology almost completely useless.
You didn't realize this was possible from scifi? What did you think the Star Trek holodeck was capable of? We saw it could make perfect recreations of characters in the show. Why would you think those characters would always be rendered with clothes on?
photoshop has existed for twentysomething years now, so the ability to manipulate photos in nefarious ways is hardly news. if someone wanted to create porn out of a family photo, they could do it with minimal skills anytime during the past decade at least. I’m really not sure why people are suddenly panicking over this.
Saw a meme that was like "one day we'll all have to say the N word to prove we're not an AI." As ridiculous as it sounds, it might come down to something similar like that. That said, one day an AI without censors will get released and we're back to square one.
In Friedman's space operas, targeting systems of space fighters have to shoot randomly to avoid being predicted (granted it's by psychics, but the theory holds)
Honestly I think with how lazy people are we are gonna find a easier way to work around this but I do really like the engagement of this video and it is making me think in a more rounded way about it
"Born too late to explore the Earth" This just isn't true, my friend - do not let this stop you from going outside to touch grass! What makes the world worth exploring isn't the piles of rocks and pools of water arbitrarily "discovered", named, and recorded, but rather the people and animals who inhabit it and make it alive; Those are constantly in flux, with hundreds of thousands of people dying every day and hundreds thousands more being born, until the skies go dark above our globe, until the oceans burn away... there will always be someone out there to discover.
@@dylanb2990 The problem is while many dislike it, many others do like it. You also have researchers who are more curious than cautious, and you have organizations with intention to cause injury or death who will continue to embrace it. LLMs are no longer the domain of large companies... see projects like GPT4All. The cat's out of the bag, the genie out of the bottle, and getting them back in isn't going to happen. Agents with malicious intent will create a demand for the tech to advance by people with intent of thwarting those with malicious intent. And the frankly better funded agents with malicious intent will just develop even stronger tech. It's an arms race at this point.
This...this really makes sense, the intro. I've left the mainstream internet for private tree houses, holes in the internet varefully curated and small. In a way, it feels like we are returning to the internet of the early 00s, where we all lived on obscure forums and chatrooms because there wasn't the infrastructure for mass connectivity.
Which was way more comfortable for me. I loath the last 10-15 years of internet. I never went to twitter or instagram, after I deleted my facebook account from school days I basically vanished from the mainstream websites (except a few select one like youtube).
What’s most dangerous is that the mainstream internet will be replaced with bots that can say whatever the corporations or politicians need them to say to give fake support or to devalue certain ideas
The fact that most of human culture says the more advanced a species gets the more they'd prey upon lesser species shows a huge reason why we can't communicate with said advanced species. Why wouldn't they become more peaceful? If you far surpass a species in technology, chances are you can do almost anything to or around them and they would be none the wiser.
Those phone scams are why I've started answering all phone calls with an accent. So far voice models struggle to replicate accents and usually mess up the voice in trying to do so, and if they do account for accent then they're using the wrong one that I feed them
Or just dont pick up anything from an unknown number. Unless its REALLY required I am not answering a single phone call. Send me a text or drop by in person. ..cant steal a voice if you cant get a connection to begin with
You could just... not answer unknown phone calls. I mean its pretty easy. only answer calls from known numbers and even then, just let it go to voicemail. if its genuinely important the person will call back.
@@shoulderpyro yea forreal. I literally never answer a phone call from a number i dont already have saved in my phone unless i am expecting a call from someone or some organization whether it be a business or other. Its just that simple. I look at people who answer random phone calls with disgust because why the fuck? If its that important theyll leave a voicemail and ill respond when i have the time.
As for voice, if you're a video/audio creator, it is advisable to only authorise banks & other businesses to interact with you in person, as noted by that guy that worked at blizzard and is a streamer & shorts producer. Then if they don't follow that notice & do something, they'll be liable because they've not followed the explicit instructions to only accept in-person instructions
I get the idea behind this, for sure, but whose bank has any goddamn idea what they sound like?? Like if you have a personal banker I guess I get it, but even for an influencer that's weird.
Pretty sure even if you do online interviews and people mess with your voice or video with AI you can still go after them because they are using your likeness without permission.
@@linkfann17next time put an at in front of his screen name. I bet he would get a kick out of seeing the occasional people saying good things about them stuff. @piratesoftware
I have refrained from staying subscribed to you for the longest time due to your over self promotional and over expectant behaviour (I hate overinflated TH-cam beggars), but I've got to hand it to you... This video is amazing, scary, and shows a much deeper thoughtfulness and depth than I had come to expect of you. I dare say... Keep it up.
Why can’t we use AI to clean out the ocean? Same reason why warehouses will be left with junk and the internet will never get scrubbed unless there’s some type of military ‘great reset.’
On thing thats important to note is that ChatGPT's ability to pass the bar exam and the MCAT does not mean it understands the law or medicine. ChatGPT can pass professional exams because it was essentially trained on the answer key. It learned the statically optimal distribution of words to spit out when given an exam question, but it doesnt actually understand the concepts, and can't apply them logically in novel situations. When a human passes the bar or MCAT it demonstrates that they understand the underlying concepts and can apply them correctly. When an LLM passes the bar or MCAT it's just a demonstration that you gave it enough practice problems to "memorize" how answers are phrased. It's the equivalent of someone passing a multiple choice test by figuring out the pattern of dots used on the scantron rather than actually reasoning through the questions
Yep. None of those AIs actually have intelligence. They don't think and they don't understand anything. It's why halucinations are impossible to fully get rid of.
You say AI "can't apply them logically in novel situations" yet that is exactly what it does, which is why it can correctly answer questions that were not in its training data. Just like the classic example, an image recognition AI can be trained on 1000 pictures of cats, then given a brand new picture it can decide if it's a cat or not. If that's not applying existing knowledge in a new situation, I don't know what is.
@@achillesa5894the thing is.. thats literally not how generative AI works. its basically iphone autofill but much more advanced, what ur saying is a drawn out version of "nuh uh"
you do not even realise that captcha was actually used to train AI... yes ALL this time you had to point out where the traffic light is, you were teaching AI what traffic lights are in different contexts
An A.I. may pass the Bar but it is currently quite far from being good at law. Also, I love being algorithmically incoherent. My wandering and seemingly pointless shopping habits, coupled with using a couple of different languages for searches has really confused my adsense. The other day I got a Spanish language Temu ad for like old lady jewelry. There's a great book called "Feed" which is sci-fi, but it has great tips on being internet random.
Yeah. On one hand it's really intelligent, on the other it has some really obvious foibles. Chat GPT needs a lot of curation to make sensible content from it.
what is considered "currently quite far from" in AI terms could be a mere weeks or months away. We need to stop thinking in human terms of references, to instead look at exponentially scalable artificial evolution. What LLMs can do today vs just half a year ago is a wider step than what LLMs half a year ago could do vs when they were first introduced to the public. Think about that, then think about this being a continuously exponential curve - at least until some hard limiter is hit, but there isn't to our current knowledge any signs of such a limit showing up anytime soon.
When I first had access to the social media as a teenager (Like Facebook & Tumblr) I thought it was the coolest thing every but after a week I felt like these media were wrong some how and disconnected myself from them and ended in small gated communities like discord and just mind my own business, these days I'm slow opening my "vault door" to these massive wild lands seeing who and what I find and what I find is pretty wild
He'd be mortified because that whole prohibition was predicted on avoiding a situation where a small handful of people controlled the masses using AI - and later, from the Butlerian Jihad, the concept of it being tabu to rely on machines to take away your mental faculties, was added to this to create the ultimate sin. And as usual, we waltzed right into it.
As an AI language model, I want to thank you for the suggestions on how to pass a reverse Turing test. No one has quite as much rizz as Kyle Hill is an internet personality and science TH-camr best known for his series Half Life Histories, which focuses on radioactive historical events.
There is a massive issue with that in person registration idea, in that it automatically restricts people who for one reason or another cannot access these locations, further preventing internet access in rural or poor areas, as well as countries with extremely restricted internets, like China
They can hire a surrogate to act on their behalf. There will likely be a marketplace of willing humans who put themselves up for hire to do meat space things on behalf of AIs as depicted in the 2013 movie, "Her"
@@bornachTwo humans being paid to meet in the real world because two AIs had to legally “prove” that they’re human in order to get something done is such a painfully believable dystopia.
Doesn't matter anyways. All that does is open up the space for "humanity mules" people who will go and vouch that they are humans but are really there to sign up on behalf of the fucking robots instead.
Aside from the scary/depressing subject-matter, I just gotta say, Kyle's 'Discount Thor' look goes great with that kind of 'firelit darkness' aesthetic. Kind of makes you feel like you're sharing a campfire with him in some Norwegian pine-forest, sharing stories and mead as the flickering firelight keeps the wolves at bay...
The fun part starts when every person you talk to on Discord can be a potential bot. They can play games with you, quote movies, talk about their lives, hold a video chat, and all this time they never had a tangible physical body. But they certainly have beliefs, beliefs that reinforce your beliefs, beliefs that the bot make see as its own but are in fact the beliefs created by its creators.
It's basically the ultimate way to scam people. Take all the effort of getting to know someone and earn their trust and just relinquish it to a hyper realistic AI. Then you have them hook line and sinker for any scam you want to attempt. It's terrifying, plus we already have people mimicking voices with AI to essentially hold fake ransoms and email elderly family members. AI is just going to take all the most insidious parts of humanity and crank it up to 11, because people will have the tools to facilitate it
So basically not that different from a real human psychology. So what? Right now people watch that video and also form their beliefs the same way. Does it mean it's 100% true? No. But we still do it.
@@cryrustmusic That's the scary part: sufficiently advanced simulation is indistinguishable from reality, and so at a certain point brains are but a grey shade away.
I retreated to a "digital bunker" years ago. The comment bots, the ads, the manipulation, the synthetic content, the hatred, the anger engagement. I hated all of it. I left the Dark Forest. I'm far happier in private servers and actually talking to people in person. It feels much more real, and like a return to almost what the internet used to be like in its earlier days.
I've long advocated for a return to the small to medium communities where we had actual moderation. Then the Kurzgesagt channel made a video about how that probably feels better because that's kind of how early human societies were: just a few hundred people huddled together for mutual support.
If going outside is the best way for humans to "signal", wouldn't the AI's first move be to make going outside undesirable? Either through making the digital space 24/7 bliss, or by making the outside inhospitable?
What scares me more than AI spamming everything is how many people apparantly take information at face value. Its always been necessary to vet whether or not something is true.
This. Yes.
Honestly, a large portion of the human population has ALWAYS lacked critical thinking, this is not new, it is simply much more easily noticeable since everyone has a smartphone in their pocket. So anyone can Google something and just spout the first thing they see that aligns with their cognitive bias, word for word.
Back before the internet, people were still doing this same thing, but were instead racing each other to libraries or centers of information to look up something that aligns with their cognitive bias.
What many people lack is critical thinking and this can most assuredly be attributed to education. By creating a universal education system online that prioritizes and teaches critical thinking skills, more people can learn to have a healthy amount of skepticism.
Modern propaganda is just repeating the same thing. That's why all the tech companies being buddies with govt is sketchy
we've been lied to all of our lives, since the beginning of organised society
So, even if it wasn't true, at least it was human. At least you were being lied to by a person
I think the fact that you might be talking to a machine is what ruins people's trust
@@goddepersonno3782there’s a difference between someone lying to your face and thousands of bots spreading false information to millions
There's a poetic justice to the idea that the reason the universe is silent is because the aliens are all afraid that if they speak up a galactic digital overmind is going to start spamming them with messages about how you can get a bigger glubark with this one weird trick.
Ooh, I just got an interesting story idea! There's a rogue super AI whose sole purpose is to learn and communicate with another species. However, it's hyper efficient to the point that it floods all communication systems in an attempt to mimic and verify the model's accuracy (not considering time perception differences). This renders a civilization's technology useless because no piece of tech can actually understand what's being said. Ignoring stuff like faraday cages (or not?), it would be interesting to see how a society develops workarounds.
Or maybe some stupid alien species did something very similar AND programmed the AI to replicate itself onto that civ's systems. Its sole goal is then to farm the species' creativity while limiting technological growth.
@@yerpderp6800 who's to say it hasn't already happened? Would we know? Do ants know they're in an ant farm?
@yerpderp6800 this was generated on chatgbt huh?
Nice DBZA reference...
Click here to chat with hot singles in your galaxy!
"Shitposting its way to the apocalypse" is a great way to describe it and should be adapted as a shirt.
Meme that. Proved my humanity 😅
@@kellymoore5517I think memes prove the opposite though. They’re anti-intelligence. Like ppl creating infinite car bumper stickers with a catchy phrase.
@@shitmandood No, good memes are distilled intelligence.
@@shitmandoodNot necessarily anti-intelligence in the slightest, just shorthands and simplifications of concepts. I mean, we have this pretty sick meme called "science" going on...
@@shitmandood more like one bumper sticker and they all say "me chad, you soyjak"
What’s alarming to me as well is the people who should be writing legislation to regulate this (or attempt to, anyways) have no idea how the internet actually works, what AI is, or what AI is capable of.
Thats why there are supposed to be experts advising them. Theres a whole army behind the scenes of analysts/ subject matter experts/ etc. but the policy process is definitely slow
No one understands what AI is - if you think you do, you haven't looked into it - to say folks don't understand AI shouldn't even be a statement.
The US is stuck in their culture war, it benefits corporations that can push their products (like AI) unimpeded
"The Internet is not a big truck. It's a series of tubes."
@@jnyboy28doesn't help that "advice" takes backseat to "contributions"
"And every article reads worse than a high schooler's first essay."
Finally, somebody said it
To be fair... its because people ARE getting dumber. We've been increasing the mount of "outsourcing thinking itself" as tech has advanced... all of this BEFORE LLMs popped off.
If only it wasn't by design so that humanity can be dumbed down to a level that makes them easy to enslave and control.... Oh well.
You forgot a period.
Even when written by humans.
@@marcosd4381it means that they’re a human, ya bot
For me the worse "articles" can't even call them there is just ones that report what someone said but gives no context of what, when or if what is being said is true or not, I wish news articles should at least give more context than just parrot what is being said to them without a thought, worse when they are quoting someone clearly lying even with video evidence in the same article showing what they are saying is completely false.
In my country a book is being banned from middle schools because it has a sex scene, which I think is weird because many middle school books here have explicit sex scenes and one I read as a teen also had an explicit rape scene, and these books never received any "Think of the children!!" arguments to be removed before, to make things worse the politicians that are banning said they never read the book they just were informated that the book had pornographic content and wanted to ban it, and what the news articles are doing? Just quoting these politicians that never read and now many people believe is full porno and I always when people start talking about it I ask what passage they think was too strong for middle schoolers and they never know, they just say that most be true that it should be banned because many people are saying its bad...
And this treatment go to anything is even for health, like many news sites spread fake news because they just quote a person and blast it in their front pages, like had people stopping their cancer treatment because they decided to quote a priest saying that the best remedy for cancer was tea.
I cannot overstate how painfully I miss the old internet...
If I would have known it would become this current dystopian hellhole, I would have never taken the old internet for granted and enjoyed every second of it.
@@dividedstatesofamerica2520Same. But I do cherish those good memories. It's really bittersweet, but it makes me feel good when I think of the mid 2000's
@@crepooscul Can't say I have much memories.
Ah the days of yahoo chat, AOL instant messenger. Making friends with strangers half a world away.
Even though it essentially started it all, even My space was good compared to nowadays 😢
Kyle recommending we use memes to combat AI overlords is somehow making MGS2 even more prevalent over 20 years later.
don't let your memes be dreams
Memes Jack, The DNA of the Soul
You right 😂 now we need a metal gear to start running around, then we have the beginnings of a reboot but in irl
I love Kyle. But he is Elon Deranged. Elon has been warning us of AI. But Elon man bad.
@@BeansMcGriddle Movies like Terminator were warning us about AI 40+ years ago. Doesn't take a fucking genius to see how bad this can go 😂
I'm not afraid of an AI that passes the Turing test. I'm terrified of one that fails intentionally.
So, are you a human or an AI?
@@codiede6690 I am a meat popsicle.
wanna see MY meat popsicle?
wait that sounded wrong
😶🌫️
Ai is the best.
As a security researcher, Dark Forest of the Internet is the best way I've heard so far to describe the kinds of ubiquitous threats the average user faces online. I'm definitely going to borrow this for educational purposes and credit this video!
Right? The 'Dark Forest' analogy really hits the nail on the head, doesn't it? It perfectly captures that vibe of the unknown lurking around every corner on the internet. Using it for educational purposes sounds like a solid plan. It's a stark, visual way to highlight the importance of cybersecurity to people who might not realize the extent of the threats out there. Maybe it could help demystify some of the more complex aspects of what we're up against online. I'm curious, though, how do you plan on incorporating it? There's so much potential to expand on that analogy, especially with how rapidly the threat landscape evolves.
-This reply written by ChatGPT4
Or you could credit the creator of the Dark Forest theory.
@@elliotgillumPerhaps, but Kyle’s video is far more accessible and comprehensive, and Kyle also clearly states the theory is by Yancey Strickler. He’s fine. Also did you know Strickler is the CEO of Kickstarter?! I didn’t!
That's what an AI would say...... I'm onto you... 🧐
The three body problem's 2nd book. Look it up, everyone here would probably love it.
"I think therefore I am" no longer cuts it. "I shitpost, therefore I am" is the new standard.
Bruh momento
Edit: shit... I just realized that future AIs will use this comment as training, so we can't use this expression anymore 😢
(You)re mad, therefore I am.
Just talking about it means it will definitely be used against us.
The T-800 hangs up the phone. "Your foster parents are dead."
@@chance9512i won a civ game, therefore, i am
@@titiwa632As funny as this joke is, the purpose of René Descartes’s “I think, therefore I am,” is for someone to prove their own existence for themselves because the fact that they are thinking at all means they must exist.
It is part of a philosophical argument to determine if everything we believe exists actually exists, or is just some illusion, but “I think, therefore I am” doesn’t prove someone’s own existence to anyone other than themselves, so it can’t really be used as a pass to prove someone is human, AI, or even exists at all to another individual.
This message is AI generated. 🙂
There is something we did on our discord server. During onboarding we have mandatory multiple choice option before anybody can use the server. The first option says "im a bot, please kick me as soon as I join (dont choose this role if you actually want to join the server)" and it actually weeds out a lot of bots joining.
100,000 IQ
whats your discord server
Brilliant infinite IQ play
Got the GOAT out here lookin for Synths lol
maybe i’ll combine this with my pun test. also, do bots try to join discord servers? what do they do?
I've noticed that when I research something, there are hardly any results anymore. The results just stop. Very alarming indeed
They are erasing in your face human logic.
Use google scholar
Same. Articles/studies I read 5-10 years ago and used for reference in discussion/debates/informing etc aren't showing up anymore. I don't think they've been erased, rather buried under curated info by corpos with agendas and adding AI into the mix is going to muddy truth tenfold.
Kinda glad when this shit really hits, I'll be on my way out of the eventual dystopia.
thats why me and a group of people are trying to gather as much information as possible and put into our "archives" a AI free community where you can find all the original information. This isn't an ad or something like that, it is a project of the future, or better said...for the future
I've noticed this too. I'm often wondering how there can be so little information on certain topics.
“Once, men gave their thinking over to machines in the hopes that it would set them free. But it only allowed other men with machines to enslave them.”
That's strong stuff
FUCK THAT'S GOOD
@@powerdude_dk attribution: Frank Herbert - Dune
I hate it when people don't attribute their copy pasta.
Frank Herbert's dune for those wondering
The irony with Captchas is that they’re literally used to train AI models.
And humana Invented that crap not ai.
Irony is that people that ever solved any captcha can nie loose Jobs due to that. Butterfly effect in action.
Captchas only look at key strokes anyways not even the images so it doesn’t matter
@@filmpeople1051they also look at key presses, mouse movements, etc. but the image is part of it
I hope the lawsuits comes as fast so these AI models get stopped as soon as possible
"shitposting its way through the Apocalypse" is too fucking relatable.
Sh*tposting is life.
It's about the only entertainment we have left now our governments have basically turned us into the culture of demolition man lol
Meme Wars II
The only aforism id actually consider as a tattoo
Bro I literally make a living out of shitposting lmao
Covid was a trial run of this
I'm beginning to understand why when my grandparents were dying, they didn't really seem to mind. The world they grew up in and loved was already gone.
Bingo. Late 21st century humans will read about the last, 20th, century like we did the Middle Ages.
Absolutely spot on. 🙋🏻♀️🇨🇦👍🥲At 62, I am at glancing in the distance at that exit sign thinking just that. Where did my world go? What world am I in even? Sometimes feeling invisible even feels comforting.🤷🏻♀️🐱
I am 59 years old and already feel like I no longer recognize the world.
@geoffreylaxton671 Sorry to say, that I completely understand. Somewhere around 2012, it seems like I entered into some horrible, alternate universe, and I can't go back home.
@@Dawn_LR What do you miss from your time? (serious question)
The "dead internet" wasn't a conspiracy theory, it was a prophecy.
Kojima predicted this in 2001
How do we know you're not just an AI competing for top comment? How do you know that I'm not an AI? I will bibbly bopino the wambly trunator, that's how.
Well, I won't be a sore loser. It was fun while it lasted. Good game!
I hear you man baby. Derp
This is why I live most of my life off the internet
I must be getting old because "back in my day" we never took anything on the internet seriously. Rule 1 was "dont believe everything you see on the internet" and rule 2 was "never tell someone your name or address on the internet". I dont know why these rules were ever forgotten in the first place.
They weren't forgotten...
They're being ignored 🫠👀
Rule 1 used to be there are no girls on the internet
Funny. "...must be getting old..." In my experience it was my parents generation who tended to believe and forward any scary email they read. I stopped counting how many of my replies to their emails contained links to a Snopes debunking article
@@bornachexcept your parents aren't who he's referring to. Before the iPhone, before minecraft, before even internet gaming, there was a generation of users small enough to have a meaningful society.
They weren't. People were gullible back then and so far as I can just as many people remain just as gullible who fall for that shit vs people who don't fall for it. And no one gives out their name or address more than previously either not to the point its noticeable. People were gullible and people will remain gullible.
Its definitely you getting old and looking at the internet with rose tinted glasses nothing has really changed on those two fronts at all.
But also its weird to say no one took anything on the internet seriously but also took it dead serious enough to never share personal information. Sounds like taking it seriously to me lol
My grandma was nearly scammed with a synthesized voice of her grandson (my cousin) asking for money because he was in a car accident. Luckily she actually called him on his personal phone to make sure and she wasn't scammed.
Whoa :/ I’m so sorry that happened. How did they get his voice? Was he a TH-camr or was he very active on social media?
I’m asking because I’m considering deleting all of my content
@@JoshuaBenitezNewOrleans I have no idea. I don't really keep track of him on social media.
@@JoshuaBenitezNewOrleans they need just a little bit of your voice. could be facebook. youtube or they called someone pretending to be some company and recorded the voice that way
I'm so sorry your grandma had to deal with that and I'm glad she didn't fall for it. We've decided on a family passphrase- something not recorded/written anywhere - that no-one outside of us would know which will hopefully avoid this situation. I think we're all going to have to do that with our older loved ones now :(
I remember an anecdote I was told during an information systems class “that if you are lost in a dark forest, the only way out may be through burning your way to safety.”
Ok, what would that mean in context of being on the internet? Crashing it?
@@pkmntrainerred4247EMP
@@pkmntrainerred4247 in this case feeding ai data that breaks it
@@pkmntrainerred4247 I think it would be more along the lines of not using any sites that you aren't sure if the content and people are real or not. Keep going; deleting accounts, until you're back to where you're sure these are real people, thoughts, and ideas, or the outside of the dark forest.
I think the dark forest is an individual psychological experience, so burning it wouldn't be something that affects the actual internet itself, but how you interact with it
@@pkmntrainerred4247 yes but that's stupid
Back in the early teens I got a phone call from some jerk. First, the call was very statically, but, and more importantly, this jerk claimed he was my grandson. I told him he was not. He insisted. I again told him he was not my grandson. "Yes I am!" he said, claiming to be in jail in California. I told him he wasn't meowing so I know he wasn't my grandson. Because I never had children, I had, and have, cats. This was not unlike the time I got a call claiming he was from the IRS, that I owed money and if I didn't him RIGHT NOW the police were going to be at my door. Well, first I had to pull over because I was in my car, and second, I knew I didn't owe any money to the IRS. First I challenged him as to what my address was. Needless to say, he could not tell me. But he kept pushing, so I told him he'd have to talk to my accountant, who also happened to be my brother. And I hung up because I wanted to go home more than I wanted to give him a hard time. I'm 70 and I just don't understand how and why people fall for this crap.
I think they’re trying to target people in the early stages of dementia :(
The con strategy is unconvincing by design-scammers want to weed out those who can catch the scam early on. Those who fall for the first scam are then targeted for more scamming.
So if you haven’t fallen for their scams, then you’ve been weeded out of their potential victim pool for now.
Lack of critical thinking skills, I guess
That's the whole point. They purposefully make their scams stupid so that only the truly uninformed, oblivious people will fall for it. Because these are the people least likely to fight them or be able to track them down and get their money back.
This reads like an internet etiquette comment
I got the "We'll have to send the police to your address unless..." call, and I said, "Go ahead."
More shouting, cursing, the Indian accent of "Agent Johnson" got thicker and thicker...
I remember like 15 years ago, Tom Scott predicted that Privacy would become an outdated virtue, and Transparency would take its place. I think that was very prescient of him.
What doesnt prescient mean
In an age where world leaders are telling you that privacy is a problem
@@unsubme2157 forward seeing, or predictive. Also could mean foresight as in clairvoyant, but I think in context it just means his prediction was correct.
@Obscurite1221 Did you read his question correctly? He said -"What does NOT mean prescient?"
@@krishanSharma.69.69f The only way to define what something means is by knowing what it doesn't mean. The opposite also holds. By explaining the meaning, you also know what it doesn't mean, as they are pairwise disjoint.
Minor correction: Text based captcha are already mostly not used because they are no longer reliable. This happened a few years back and is the reason most captcha switched to just a check box or newer ones which are entirely behind the scenes.
Yeah, they basically look at how you move your mouse pointer to click the button. Apparently it's hard for a computer to copy that still.
@@BlazinLow305 Its more about just mouse movement. Its previous browser activity, cookies data, and all sorts of other information. This is why if you're in incognito or something, where cookies arent stored, you're far more likely to have to actually do more captchas after clicking on the checkbox.
Its not so much that AI "cant" copy these things, its more so that these sorts of things take a lot more time and effort to copy well, making it more expensive to bypass. Plenty of bots are capable of bypassing these captchas, but the goal is to make it as time consuming and expensive as possible to make it not as worth it for bot farms. Most bots used in bot farms are extremely simple to save costs. Scams are trying to make a profit afterall, so higher operating costs makes them less profitable.
came here to say the same thing. blurred word captchas haven't been used in YEARS. that was kinda weird to see him so out-of-date with that bit of info.
@@eragon78 yeah all captchas can be bypassed without exception. The only question is if it's worth crawling a site with a nasty protection or not.
There was a point where the way to detect a bot is that it passed the captcha too quickly
I wonder if the internet will be so terrible in 5/10 years that people will be scared to go there?
(Wouldn’t be the worst thing…)
As someone who barely uses the Internet due to self improvement and mental health reasons, I'd fucking love this
I think 2-3 years
You underestimate horny people
I appreciate how you use the same straightforward tone and quiet music that you use (presumably out of respect) when discussing people dying of nuclear accidents. At the same time, it's quite chilling.
Eh another doomsday prophet. Only talking about the downside. Or ye, he added 15 seconds of "It can also be good". Do 15 minute video of the literal utopia that is coming and I will take you seriously. If not, you are just a doomsday prophet.
oldest trick in the book playing minor keys over a sob stories.
Our parents' warnings of "don't believe everything you see on the internet" have an eerie prescience in the age of algorithmically-generated content...
It's soon going to change to "Don't believe ANYTHING you see on the internet" but we all know there'll be people who will continue to believe just about anything they read.
Ironically it's often parents i.e. older people who're much more vulnerable to online disinformation. Cos they're not in the habit of being skeptical or checking stuff up. Their security came from trusting people they know. So all it takes is the message to come from a trusted source who's been hacked and they'll accept it without question. A lot of sketchy alternative medicine ends up marketed this way for instance. Despite it being called a 'curated community' here, Whatsapp is full of viral chain mail crap.
In my experience I have to warn my parents more than they ever warned me.
@@lorpen4535 This is basically what I said here, but YT in its algorithmic wisdom seems to have deleted my comment. But it's still sending me notifications of other comments. Trillion dollar company btw.
Now I’m telling my parents this
Synthetic or not, the corporatism and cleaner internet was what truly began the beginning of the end.
From the moment someone tried to curate the internet to make it more marketable was already on the path of death.
I mean these problems are just as bad on search engines as social media. If search drove as many views and people spent the majority of their time on independent websites the bots would just make those instead. Making a website is just as easy as writing text to an AI.
@@rjohnm666 The beauty of that is that it's not always people. There's entire botnets dedicated to creating political discourse now. The arguments you hear may just be trained autofill bots, bots whom the sites on which they post will not counteract because they create engagement. It's insidious.
But it's not just that. AI content farms and bots have their own incentives to exist and have no relation to the "clean internet" or whatever.
How I miss the old internet that was like the wild west and not run by a handful of massive companies.
The only reason the internet evolved from a DOD project into something useful for civilians is the desire to make money.
There have been numerous academic papers recently that have shown that bots are more effective at solving CAPTCHAs than humans are. At least for many of the common types of CAPTCHAs that are currently in use.
Plot twist: you've been training an AI every you time you fill out a captcha
youre right and actually the captchas doesnt detect a well made performance in being completed but the opposite of that to recognize the human common behavior that is more erratical
Don't talk about how captchas are passed on a video about ai taking over and creating a dead internet. Bots will have new features in no time
Currently it's more like you must prove you're a human by failing three captchas in a row.
I'm human.
And that's why I never post photos of my kids. Family are like "why you hiding them from us ?"
Now I can show them why, with this video.
Thanks, Kyle.
There is a Dutch add that warns against posting your kids on social media and the dangers of AI with it.
You should be able to find it on yt, show them that short add and then they will never ask you again
Even 20 years ago that was the right approach. Good thing you've been going that way. I worked at a free-hosting company around 2000-2001, and we regularly provided evidence of CSE rings that got them busted, but there were also things like a family innocently posting beach day with their kids, and some creeps found it, and suddenly their kids' photos specifically got 20,000 hits overnight.
It's gotta be horrific for them realizing their kids' pics are being shared by pedos for arousal purposes.
Or because you hit them? lol
I will never post my kids online if I ever have them.
And from now on I will always put on my pictures a text bar "I do not consent to this image being used to train any AI model." If enough people start doing this, it will force the creators of AI models to stop scraping images from the internet, and force them to ask permission to use an image on their training data. Because it doesn't matter if a website says yes if its users (aka the people creating the content) are never even asked.
“Why are you hiding them from us.”
Come visit tf? You’re always welcome.. ?! Besides there are always private family photos albums that you can post to…
I can’t imagine being entitled to someone else’s only appearance.
The internet used to be the one place you could go when you were lonely - regardless of how awful you felt - and at least be connected to other humans in some form. Feeling that disappear is scary.
Hey, at least we have that to connect us.
Beep beep boop boop I am real human click link to buy butt plug
Maybe it's my age, but I think the restoration of a balance between 21st and 20th century interpersonal dynamics would be really nice. I don't think we'll totally lose the internet but I would LOVE to see the destruction of current social media culture, which I never fully embraced. All the oversharing, privacy violations, toxic trends, and shit like that, imagine the distrust being enough that one day, people just turn off TikTok. (An app I refused to get, BTW.) Some internet will still be good and useful but being pushed back outside would be a massive boon to our mental health on a societal basis.
Maybe this will cause people to get out in real life and make real friends, might be a net positive.
Ironically enough, an AI therapist is or could probably be just as good if not better at making you feel less lonely. Some people already find AI therapists to be helpful.
A buddy and I were talking. He likened it to how phone calls are a dark forest already. It’s potentially hostile to answer a call from a number you don’t know. less scammers find out you're a real human. You only call text/people in your phone book. I think the surface internet is rapidly approaching a similar dynamic. This has precedence.
Hm, haven’t thought about this aspect.
Very insightful and completely true, damn.
I never thought about it that way but that's actually true :0
I used to make a point of answering all calls because I'd rather get some spam calls than miss a single real call. But the few turned to many, the many to excessive amounts, the excessive amounts to unmanageable amounts of, and at this point not only do I trust Android's spam number detector, I don't even pick up calls from private numbers because in the last year, no exaggeration 100% of those has been spam calls. It's impossible and I've stopped trying.
I mostly don’t answer my phone if I don’t recognise the number or have it saved. It’s frustrating to near constantly get robo calls or calls from someone who’s clearly trying to scam you. Especially as a night shift worker, they wake me up with their nonsense 😒
It's not all bloody doom and gloom. Don't get paranoid or anything that'd just DUMN
I get scared when i start writing a search on google and it automaticaly auto-writes my question with what was in my head
As someone who's passed through everything from "You're the man now, dawg," to "Do you know de wey?" and on to gyatt rizzler, I gotta say I'm not convinced me or a lot of people in my age group could have any chance of catching up on the latest memes faster than data-scraping LLMs...
they may even invent the memes faster then us at this point
I don't even try anymore, what's skibidi even is or means...
@@NoahGooderAI generated predictive memes are the future
The latest brainrot memes could be the only way of human to human communication in the future, like a secret password to enter a human hideout, or a code, encryption of sorts to confirm you are talking to a human. memes change weekly and spread like wildfire. it can reach humans before it reaches ai.
AI can't shitpost the same way we do, especially with images because they cant look at it, decipher its meanings or what makes it funny and repost it sith actual feelings. So personally i doubt the bots can keep up with US when we have accounts entirely dedicated to making and reposting goofy ahh memes.
13:58 “Maybe Rizz and Gyatt will end up saving something uniquely human online” is not something I expected to ever hear in my lifetime but here we are
The future is thicc (verified)
This statement alone was one of the bleakest lines in the whole video to me.
What the heck is Rizz and Gyatt?
He needed some phrase like to prove he wasn’t a bot lol
@artugert you know 😏
Humanity is defined by starting to massively use something new that it does not still fully understand and dealing with the concequences later.
Lead, phosporus, radiation, asbestos.. those are just the ones i can come right now from the top of my head
The technological revolution and its consequences
mercury, antibiotics...
Add fossil fuels to that list. And even though we know for a fact that the continued use of fossil fuels is going to end us, we continue with business as usual, because switching to a more sustainable way of living isn't profitable right now.
trust science until you don't
What happens before one gains indecipherable for ours consciousness......
In a few years, we're going to find out AI was active a decade or two before we were told originally.
makes you think how advanced it could be right now behind closed doors
@@SteffenSkyHi Correct 💯, probably 50 years beyond the public stuff
"No LLM will be able to keep up the pace with weird internet lingo and memes". This is a funny statement, I wonder how much time will it take to age like milk.
I guess depending on whatever exactly Q* is (the new super AI that OpenAI is developing) it may be already outdated.
A certain and easy answer is when AGI is born. If one ask for my non-trivial prediction, then I'll say that it'll be so when photonic computing are used extensively.
Yup. The Equivalent of saying no Locomotive can Outspeed😊 a Racing Horse.
We already have RAG models, they may not be perfect now. But it has the potential to prove that statement wrong.
NVIDIA has already implemented into their own chat interface. Huggingfaces' chat page also has RAG capabilities.
An LLM with a long context window like Claude 3 could probably already keep up with Internet culture if it is given Internet access and runs in a loop.
"You were so busy trying to find out if you could, you never stopped to think if you should." -favorite Jurrassic Park move quote.
"Thou Shalt Not Create A Machine In The Likeness Of A Human Mind"
Nah. If generative ai wasn't profitable they wouldn't even look at it's direction
@@venomshot2815 "We're going to make a fortune with this place"
Iian malcolm would have mich to say on this topic and how we deserve whats coming
I’m on Elon musks side with this. Someone was going to create this no matter what. If it’s not good actors with good intentions then it’ll be the bad actors who develop it. It’s important to explore these things regardless of the danger because of good people find solutions to keep it from being a danger that’s a whole lot safer than bad people having access to the technology and doing anything they want with it.
I think a lot of people fail to realise that AI doesn't have to be "smart" for it do be dangerous. Much like early computer viruses, a piece of code with a simple objective in an ecosystem not ready for it, much like introducing a predator, can be catastrophic. Add to this the genuine loss of tech literacy with generations who are growing up with tech "just working" and the ability to even understand how it's a threat at all diminishes. One of the most common issues I've had to assist students in troubleshooting, at a university level, is running every program under the sun all at one and a million chrome tabs, or not understanding that their device has a physical limit on storage. this results in complaints that our software is buggy or doesn't work for them, when in reality they fundamentally misunderstand the machines they use for most of their waking lives.
I think people also fail to consider non General Intelligence as an arm of human interaction. The machine may not have wants but it's operators will. It will be the mask they put on and the megaphone they use to drown out dissenting ideas.
You could try giving them the task of protecting the Internet by giving them a box with a red light on it (Tell them THAT is the internet and they must protect it!) Ref: The IT Crowd Season 3 episode 4
Luckily, increase in focus on STEM will HOPEFULLY increase tech literacy in the younger generations.
In middle school (not many years ago), peers would ask me for the time, so I would point to the nearest clock, only for them to ask again or ask to look at my phone for the time. This level of tech illiteracy does not just extend to clicking a few buttons on a device, but also to reading analog clocks, something we were all taught in kindergarten.
I’ve been a programmer for over 17 years, one of my biggest fears is when I inadvertently fixed a bug in my code and don’t know why it’s working. AI can generate code for you now, and this is a nightmare. It’s difficult enough inheriting a code base you didn’t write yourself, imagine inheriting one that NOBODY wrote.
14:55 dude..... I have a 6-year-old daughter and that statement that you just said just really hit me hard. Thank you for that. My goodness.
Pretty sure the tiny 'thousand yard stare' after he said it, was genuine.
I came here to say this. As a father to a 4 year old girl I had to stop and swallow that lump in my throat
@@facemonkeys God bless you homie
@@ChristopherGonzalez1280 you too brother. Keep the family safe
It is truly horrific. I hope the people that create and support such abominations are caught and rightly punished.
Personally, I think the only way to combat intelligent bots is with more intelligent bots created for the greater good, so hopefully law enforcement could utilize A.I. for the better. I'm not optimistic, though. To anyone who read this, stay safe and try to keep your private life off of the internet 🫂
"When's the last time you believed any headline, any social media post not wanting to check if it's truth / has the right context"
Probably some time before I actually got a little media literacy, clickbait in media existed long before even youtube and it made blind trust in headlines a bad idea way before AI was a thing.
This completely. Propaganda and agendas are not something that was magically conjured into existence because of the internet or wasn't widespread because of it. It just made it easier to spread new versions of it.
Yes. AI lets us mass-manufacture BS, instead of relying on the handcrafted, artisinal BS that we previously enjoyed. The BS itself is nothing new; just the overwhelming quantity of cheaply made knockoff BS is.
Was gonna say, the 2016 election, and coverage of an unrelated individual, is what really woke me up to the blatant lying and deliberate misconstruing of context the legacy media is guilty of. Does not matter the ideology driving the specific organization, they're all guilty of it. At the very least, that has prepared me for what LLMs will do.
@@maxsalmon4980 All the corporate news just copy each other as is. Quantity doesn't matter when traffic is dictated by what big tech wants to spam you with.
The point still stands though. AI allows such a large amount of false information to be created so quickly and spread instantly. It's going to get to a point where you can't even check the truth or context of the original post, because every new thing you read while trying to search for the base facts will also be shrouded in fake posts and articles. And those fake articles will be so similar to real articles that it will be next to impossible to know which is which, unless you watched whatever event you're researching firsthand and don't need to look anything up in the first place. It's more than just being smart enough to not trust everything you hear, it's going to become almost impossible to even research things online at all. Our legislature, ways of detecting AI, understanding of AI, plans for how to handle AI, etc. are all lagging behind the actual proliferation and creation of new AI technologies and are only going to fall further and further behind if we don't stop to address it right now. As a Software Engineer who has taken multiple classes on AI and AI ethics, AI is a terrifying thing in its current state, and not enough people are planning for how to handle its progression.
I never would have thought ive been training my whole life to combat AI by being constantly up to date on memes.
Yeah me too 😂👍
Said an A.I to prevent you from achieving to prove your difference from it. ^^^^
Keep training bro! :D
all i know is that chat gpt will never tell me " i need to have a deep sniff of that bussy" and i cant tell if this is a good thing or a bad thing.
bot, pls
memes also taught me how to talk about capitalism, transphobia, mental health (namely, mine), and the ongoing epistemicide of indigenous people. (fresh memes also make an excellent nazicide!)
though the whole idea of there being effective methods of "combatting AI" (the possibility of boticide) is widely regarded as a pseudoscience among the bot scholar community, and generally sneered at in botiversity campuses.
As someone who has been using the internet since the early 90's and has made an entire career developing websites, I'm ready for the internet to just go away.
I have constant Carrington Event fantasies
I'm ready for the internet to only be allowed to be used as an archive for libraries. I mean, the revolutionary thing about the internet was the easy access to information, so let's make it just for that. Only libraries can put information in, and we can only use it to search and access the information provided.
@@NankitaBR Putting information in... where?
If your client computer can't upload data, you can't physically request the data you want to download and the internet becomes entirely useless.
Most internet resources already are the way you suppose - curated content that can only be consumed, not added to except by whoever holds the keys. So what you're suggesting is basically turning off comments on sites that allow user input and engagement.
At the least there's a deep irony in posting a comment to a youtube video suggesting this!
Or at least go back to the way it was back around the mid 2000's that's when it was great
Same. I basically grew up connected to the wall and at this point if I could I'd move out to the woods and join a commune.
We're one step away from the internet becoming a literal Infinite Tsukuyomi.
NGL, had to look that up, and was a little 'put off' by it being Naruto-related. Then, I read the description... 100% spot on.
Great video!
Just a note: distorted text and image recognition captchas have been automated with high reliability for a few years now.
That's why there's new techniques such as align the puzzle, find the identical figure, etc. But even those are getting automated sort of easily nowadays. It's kinda mouse/cat situation.
Exactly there is incentive to make captchas only solvable by humans and in the other side there is incentive to automate the solving, it's an arms race.
from what I understand about those tests, they're really just tracking human imperfection.
and while that can be emulated with little effort, if any solution gets used for mass automation, the method will end up tripping alarm bells and become unusable.
Have to be that guy, but those captchas are only being used to further train AIs. Captchas filter bots from people using mouse movement and cookie data.
So if captcha cant protect us, then whats the solution to avoiding A.I bots? Just dont use the internet?@9308323
biggest issue is that google is using reCaptcha as a training data source so anything they come up with to prove you're human will just be used as training data to make a DLM that is capable of solving it
google in their greed is just shooting all of us in our feet
this whole concept reminds me CONSTANTLY of the Cyberpunk (ttrpg) concept that the open internet is so garbage and virus laden that there is a whole career based around finding things for the layman
I feel like we're not as far from the need for Agents as I would like
I think we're pretty much already there sadly
Since Twitter became mainstream I was surprised the job did not exist, and since then more and more art critics and human curators were simply replaced by algorithmic recommendation systems.
R.A.B.I.D.S. is basically already among us destroying everything useful. Soon the only place with any valuable data left will be private networks...maybe we can call them data fortresses?
Exactly. It instantly reminded me of watching Cyberpunk:Edgerunners a while ago.
have only plaed the game (2077) and everytime i hear one of these news about AI all i can think of is about Bartmoss, the BlackWall and Netwatch agents
Like, why is it happening, why can't it just stay in a game as a story, not as a warning?
I have never been so attracted to the idea of moving to a cabin in the woods with no internet
Ok then Mr Kaczynski...
@@Dalroc history will completely vindicate him
@@Dalroc I don't plan on mailing bombs from there. He was clearly unwell from a young age and I.. okay not well but in a totally different way 😜
@@Dragonflyer74 he had valid concerns but the way he acted on them was fuct. Even if you thought killing random and basically blameless civilians to make a point was cool, it was obviously ineffective, which makes it stupid.
Why cabin in the woods? Plug the internet cable out and turn off mobile data. Done, no more internet for you.
7:46 there's something deeply existentially terrifying about the fact they requested a list of human traits AI can replace and it gave a response, and one of them being "empathy"
I imagine those squid robot things from the matrix swarming around desperately looking for someone to show ads to.
Terrible thought. Gave me a laugh.
@@danyael777terrible laugh. Gave me a thought
They will generate engagement, by any means necessary.
agent smith tracking down adblock users
@@NoNameAtAll2 Agent Smith got "promoted" because "he" had the best audience reach metrics.
Fun fact about captcha. The test for it starts before you reach the captcha. It uses your internet history and mouse movements to determine if you’re a bot. Bots go straight to the website with no history, move very quickly and in straight lines. So in order for a bot to pass it they need to do unnecessary movements and purposely be slow.
If you train a bot on mouse movements from a data set of humans and give it a fake history of one of said humans it could pretty easily pass the click test.
no fucking way a website has accesss to my history?
@@transsexual_computer_faery depends on your cookie settings. But ye, generally it's the mouse movements they use
That's pretty easy to get past even with a simple script. you just chuck a few random variables and pauses in and it no longer looks predicable.
@@asmosisyup2557 eh, not that easy. The variables and random directions are always mathematically perfect. To truly get the simulated randomness, you would honestly be better off recording a human's mouse movements and using those.
This is such a distressing concept for me. I already feel disconnected from everything and everyone in real life and being surrounded by an internet of computer generated content is going to drive me insane.
Take up a hobby that doesn't involve being online, like playing a musical instrument, knitting, model-building, etc...
These activities can be quite stimulating to the mind without the constant barrage of bots.
As a generative AI model, I think everything will turn out fine.
I'm just kidding, I hope you're not in too much distress over this stuff. I think the comment above is pretty important, thought. Taking up an offline hobby, specially one that makes you gather with people who share the interest, like chess or a trading card game, might do wonders at a time like this.
Probably time to start reading again, huh? The internet as we know it is about to die.
Maybe it will be a good thing. I see this to have the potential to ruin the internet as a place to just exist in as a surrogate for real human interaction. Maybe we'll all go back outside again
Have you considered going out of doors?
For decades I’ve used the internet to fill the social gaps I’ve experienced in life. I even met my wife online.
But this video brings to the forefront of my mind something I’ve been feeling for a while but haven’t been able to place. The lack of humanity present on the modern internet.
It’s a strange thought, that I may be of the only generation that actually found humanity online. Will my children understand that some of my strongest moments of bonding with others occurred in a forum?
We’re entering the internet 2.0, and I’m not sure I’m ready for it.
This would be 3.0. 2 was social media
Me too! 😢
Bro think he shakespeare
@@shorx9199 only if your dumb
I feel this the internet and gaming is the only thing that keeps me alive and safe from depression if i lose it then i want to die
I love the fact that you are telling this like a story from the past. Like something you will tell your grandsons in the woods while camping.
he's talking to us from the future, when the last remaining humans live out there in the woods while the cities are populated by different AI interacting with each other
True, all of the ideas are already happening, they're not even speculation. He's playing catch up at best.
The internet has been dead for a while. With all the "SEO Pollution" as I like to call it, Google and etc services have been burying many things people are actually looking for. Not only that but Google also nowadays makes it really hard or impossible to find really old images and etc. See "The Dead Internet Theory" for more on this subject and then consider when this event took or will take place...
*grandkids 🙄
lol, do you mean glamping or is it just a glamorous way of looking at how your reality will become? I mean, not camping of course, but living in the woods, since we're now considered trash and don't deserve living in a city.
Those statistics about the amount of generative content are absolutely horrifying.
We are definitely not ready for whats coming.
The stats were generated. :)
whats horrifying is that curated spaces will not be capable of filtering bots in a few years. Imagine advertising bots that spend years talking to you to gain your trust, capable of joining and talking in VCs and playing games and sending consistent essentially photorealistic selfies.
just to add to this, we already have rudimentary systems that can link multiple models together to do what chatgpt, dalle, and other models did together. Conceptually this already exists, its just not yet good enough to fool most people.
@@badabing3391 So a Cylon as a friend? .. beats having to be social with a human
to be fair its mostly capitalism. if we had no reason to be afraid of corporations trying to use humans as capital, we basically would have no reason to be afraid of AI. if anything we would be happy with automation replacing labor so people would be free to pursue things that give their lives meaning.
“Thus did man become the architect of his own demise.” -Second Renaissance, Animatrix.
Just smart enough to be the architect of his own destruction.
oh dayum i didnt expected to see a quote from Animatrix, holy moly
"Man created machine is his own image." God, I think about the Second Renaissance so much.
"Disillusioned words like bullets bark
As human gods aim for their mark
Make everything from toy guns that sparkTo flesh-colored Christs that glow in the dark
It’s easy to see without looking too far That not much is really sacred" Bob Dylan 1965 - This verse hits home for today's time.
What happened tto the Animatrix? It was so good but it is hardly seen or referenced at all.
Kinda hit home with how i won't respond to random messages unless i actually know who is messaging me. One thing I've noticed about AI is that it isn't very good at saying it doesn't know an answer to something. It always tries to answer.
yet.
Fortunately, human beings have a great track record of thinking carefully about our next steps.
🥲
Yah straight into a hole, this time an abyss.
biggest cap ever
Sarcasm right?….. right?
lololololololololo
Outstanding analysis. The real upbeat ending here is that we can all put down our phones. When I play terraforming mars with my friends and go to ace hardware and then work in my yard to build raised beds for my bulbs, it's all people... and there is no doubt.
Terraforming Mars is so fun! One of my favorites
Let's practice algorithmic incoherence...
Regarding TMars.
Blue's early, build engine.
Prelude? So many fiddly bits!
3D printed boards for xmas: so much WIN!
Got Turmoil. Hella complexity spike.
Stop driving cars, riding trains/buses and planes while you're at it.
It's like we're creating two worlds that are growing apart. It reminds me of cyberpunk
@@dividedstatesofamerica2520what does that have to do with anything
It certainly feels strange to live right now. We have just witnessed the explosion of something as impactful as the atomic bomb in just a few months, and we still don't know how to react
Yep. We let an airplane take off without having a place for it to land.
Clever.
I used the same analogy when I wrote a paper about this 10 years ago in college. Not to sound self-important, but these past ~13 years for me have been like being tied to train tracks and watching the engine thunder towards us. There's nothing I could ever do about it, but I had to watch it coming.
This is the first blare of that awful horn, and any who weren't looking before, surely know it's coming now.
@@dan_kay It's more like, we are all on an airplane that turned out to be a rocket ship and we are moving into the blackness of space.
@@dan_kay"we let"?
Who exactly asked for this AI stuff?
@@Blackfox_o7 it's more like we are living a bunch of dystopian sci-fi movies all at once. Does art imitate life or does life imitate art?
I love this one. Plus, the emphasis on getting outside, it seems you filmed this in the evening next to a fire. :D
As Metal Gear Rising Revengeance once said:
MEMES, THE DNA OF THE SOUL
I swear, that game was borderline prophetic.
@@DoctorLoudonclearfr
@@DoctorLoudonclear this applies to almost all the Metal Gear games.
I wonder if AI will cause our language to evolve into something similar to that world from TNG: Darmok. Everything we say communicated in referencial memes.
Well, that is pretty much literally what Dawkins wrote when he coined the term meme in The Selfish Gene
Once the internet is overrun by AI, we'll be forced outside our homes and meeting each other in person. Its kind of a self limiting problem.
A comforting thought, actually.
What do you mean "once" the internet is already infested with AI just do a google search and the first site will be ai generated and full to brim with ads.
I hope, but people have gotten used to advertisements being everywhere, I think people will get used to AI
People still meet each other outside. Only people like you and me are trapped in this.
Perhaps in regards to socialization and the exchange of ideas, but with every corp business practically forcing you these days to use their website or app, that opens the door to any number of issues. And never underestimate the amount of ridiculous treatment people will accept towards themselves (or others) in the name of status quo comfort.
I feel like that this will force two potential of many outcomes to occur:
1. The internet gets scrubbed at some point of the influences of A.I. somehow
2. A new "socialnet" or something gets created separately from the internet where only verified humans are allowed to interact with it, and it is ruthlessly guarded against A.I. content regardless of its origin.
3. Everyone back at the pub. I vote for that.
The problem with these to outcomes is the same: who does the policing and how? And more importantly, who monitors the monitors?
@@daikansanchez7674We could probably script an AI to monitor it. It has the advantage of being impartial.
@@cocolasticot9027No, we don’t need more drunks in this world
You can't prevent the use of AI in your new socialnet... How can you prevent someone from generating text and just coying and pasting it in a page on your socialnet. The genie will not go back into the bottle.
We never would've guessed we got so deep into the internet that we're starting to get pushed out because of it
The irony of the internet being the place that would bring us together, devolving into an isolating cacophony of meaningless content, is quite something.
I love how you worded this.
Yep… quite a turn of events… I’m curious if any of the great sci fi authors got this right I can’t think of any but I’m not sure
So has every communication medium before it
It's the circle of life.
For the power of money.
In the circle.
The circle of life.
Fate, it seems, is not without a sense of irony.
This morning was an eye opener. First a post from an artists complaining about shite AI generated art instruction books, then my wife telling me about AI generated crochet patterns that can't be made in real life, and now Science Thor warning of the dark forest. Gah! Even the algorithm is involved! I'm going outside with my dog.
how do you know your dog isn't an ai?
@@omatic_opulis9876How do you know that @Naturallystated isn't an AI. How do you know that I'm not an AI?
@@mateusvin well if i ask what did Kyle say in a specific part of the video and you answer that could work right?
like what did Kyle say/do/show at 12:00? bots would have a hard time with that
@@mateusvin as an ai language model trained by openai, i can not provide you with comments or replies.
@@radwanshakfah6938have you not read up on the latest Gemini AI? It can effectively read a book, or watch a movie and recall info about it.
This is why I'm stuck watching a live cam feed of eagles hatching in a tree top.
North East Florida Eagle Cam? that is the one I watch and it's so cute
What if... That's AI generated and you don't even know it because that AI is just THAT good?
Eagles aren’t real. They were generated five minutes ago and your memories aren’t real. Lol
birds in a field for me...but ARE THOSE EVEN REAL NOW????
@@shalimarsnow9316 I think it was Morpheus that asked, “what is real?” Electrical signals processed by your brain.
I mean if we are all AI it doesn’t make our existence any less important. Real is whatever we make real. We still can understand when something doesn’t feel “right” like hurting another person or telling a lie.
I don’t know. There’s still something to us, whatever we are.
Kyle I just want to give you a huge compliment and virual hand shake as I absolutely love your content. Utterly engaging, fascinating and educational. This alone has opened my eyes and honestly I found it difficult to comprehend some aspects of what AI is becoming, but thank goodness you explained it the way you did! Many thanks to you from Melbourne Australia :) - and yes I am a real human btw !
Captcha doesn't just rely on reading and reproducing distorted images of text. Some captchas are just clicking a checkmark. They actually track your mouse movements, browsing history, etc. to determine if you are a bot.
That was ReCaptcha which is a similar model but different company. RC doesn't use distorted lettering but instead uses directed images.
we also have these things and are very much trying to implement them to allow for data verification. Provenance isn't exactly a big deal on the bigger scale with c2pa and friends - but all that is sort of moot when you consider the current climate of nonsense: lobbying and anti-journalistic sentiment is already so high, all gen-AI is going to do is level the playing field, at least somewhat. Plus: who says I want to explicitly point out my humanness online - or talk to other humans for that matter? I mean, 99 % of people who have used an LLM to bootstrap their learning endeavors can perfectly put into words why they prefer machines over the real deal. Even great teachers can be real pricks at times, dealing with an individualized chatbot tailored to your specific needs, one that won't make you feel ashamed for asking easy or silly questions... that alone is going to completely revolutionize very "RL"-type of tasks. The Internet already was just platforms allowing you to curate specific feeds, reddit is literally called the frontpage of the Internet, and face it, it's not like the big subs could get any worse with synthetic content. Which you can still moderate, catching vague shilling or anti- sentiments isn't actually rocket surgery at this point. If all fails, new communities will crop up, but honestly, chances are we'll just enjoy talking to synthetic communities more. It already is way better than exposing yourself to Stack Overflow, among the most judgmental communities next to German pinboard culture.
When the discussion came up for how to confuse AI with cultural lingo and memes, it made me think of Kevin's "small talk" from the office. 'Why waste time say lot words, when few words do trick?'
Reminded me of the ST:TNG episode "Darmok" in which Picard and crew encounters an alien race who communicated entirely via meme which rendered their universal translator technology almost completely useless.
@@bornach Arnock, on the night of his joining.
@@bornach The "funny" thing is I am not sure English is any different from that. Kind had to suspend my disbelief for that aspect.
Why time waste words? Few do trick
@@Mantosasto Temba, his arms wide
Kyle slowly becoming the 4th Vsauce channel.
This show is much different and I like it so take that Vsauce, wherever you might be
Nah, Vsauce sold out, Kyle is the real deal.
@@michrain5872 who did he sell out to?
I miss Kyle from 3 years ago. The new quiet voice, aka "Serious Kyle" isn't as fun or engaging as Quiet Kyle.
Where's my Kyyyhhel?
Kyle hill > Vsauce 🤷🏽
As much as we try to see A. I. in a positive light, , opening new doors etc., we all know this is going to end in a really distorted way
"Pornography of your daughter from a single photo" hit home exactly the way you intended it to lol
"Nudify apps" Yep, I see where you're going. "...your 10-year-old daughter..."
That...that hit like a ton of bricks.
Such a fked up thought. That absolutely will happen to tons of people. 😢
You didn't realize this was possible from scifi? What did you think the Star Trek holodeck was capable of? We saw it could make perfect recreations of characters in the show. Why would you think those characters would always be rendered with clothes on?
photoshop has existed for twentysomething years now, so the ability to manipulate photos in nefarious ways is hardly news. if someone wanted to create porn out of a family photo, they could do it with minimal skills anytime during the past decade at least. I’m really not sure why people are suddenly panicking over this.
@@mencibenci That's like comparing an 8 year olds flipbook animation to 4k video
Best way to make sure someone is a human on the internet is to ask them to swear or to send you something protected by copyright.
For now, but uncensored models are easy to find and peoples who want to span or scam other users won't bother with censored AIs like chat gpt.
Or ask for conspiracy info on their producer
Swearing isn't a realistic litmus test, chat bans are too common now.
That will only work until someone manages to force a future generative AI to write code for generative AI without those restrictions.
Saw a meme that was like "one day we'll all have to say the N word to prove we're not an AI."
As ridiculous as it sounds, it might come down to something similar like that. That said, one day an AI without censors will get released and we're back to square one.
Once again, Star Trek called it. Theres more than one episode where the Crew defeats some AI or Android menace by being illogical
peace and long life 🖖
@@driftingdruid live long, and prosper 🖐️
As an AI model, I wish you safety and a long existence.
But what if someone makes an illogical AI? How to defeat that?
In Friedman's space operas, targeting systems of space fighters have to shoot randomly to avoid being predicted (granted it's by psychics, but the theory holds)
Honestly I think with how lazy people are we are gonna find a easier way to work around this but I do really like the engagement of this video and it is making me think in a more rounded way about it
Born too late to explore the Earth
Born too early to explore space
Born just in time for existential crisis
Thanks for the nightmare fuel man :P
It's so much worse than you think :,)
Be ready, this is the point of no return.
"Born too late to explore the Earth"
This just isn't true, my friend - do not let this stop you from going outside to touch grass!
What makes the world worth exploring isn't the piles of rocks and pools of water arbitrarily "discovered", named, and recorded, but rather the people and animals who inhabit it and make it alive; Those are constantly in flux, with hundreds of thousands of people dying every day and hundreds thousands more being born, until the skies go dark above our globe, until the oceans burn away... there will always be someone out there to discover.
@@Ilamareait’s really not, everyone dislikes where it’s going, so it won’t go that way. This is a temporary problem.
@@dylanb2990 The problem is while many dislike it, many others do like it. You also have researchers who are more curious than cautious, and you have organizations with intention to cause injury or death who will continue to embrace it. LLMs are no longer the domain of large companies... see projects like GPT4All. The cat's out of the bag, the genie out of the bottle, and getting them back in isn't going to happen. Agents with malicious intent will create a demand for the tech to advance by people with intent of thwarting those with malicious intent. And the frankly better funded agents with malicious intent will just develop even stronger tech. It's an arms race at this point.
This...this really makes sense, the intro. I've left the mainstream internet for private tree houses, holes in the internet varefully curated and small. In a way, it feels like we are returning to the internet of the early 00s, where we all lived on obscure forums and chatrooms because there wasn't the infrastructure for mass connectivity.
This is largely the fault of something Corey Doctorow calls enshittification. Predominantly driven by social media.
But isn't that the definition of living in a bubble, which comes with its own set of risks?
From my point of view, forums and chatrooms already were mass connectivity.
Which was way more comfortable for me. I loath the last 10-15 years of internet. I never went to twitter or instagram, after I deleted my facebook account from school days I basically vanished from the mainstream websites (except a few select one like youtube).
What’s most dangerous is that the mainstream internet will be replaced with bots that can say whatever the corporations or politicians need them to say to give fake support or to devalue certain ideas
What needs to be remembered, is that amazing outcomes aren't always positive.
For anyone who's worried: I highly recommend looking into "PauseAI". There are things you can actually do to help.
All technology is inevitable. Call it pandora's box, the tragedy of the commons... We have no real control, it's our fate to go along for the ride.
The fact that most of human culture says the more advanced a species gets the more they'd prey upon lesser species shows a huge reason why we can't communicate with said advanced species.
Why wouldn't they become more peaceful? If you far surpass a species in technology, chances are you can do almost anything to or around them and they would be none the wiser.
Those phone scams are why I've started answering all phone calls with an accent. So far voice models struggle to replicate accents and usually mess up the voice in trying to do so, and if they do account for accent then they're using the wrong one that I feed them
Noted.. good idea
Thank you for your wisdom Knuckes
Or just dont pick up anything from an unknown number. Unless its REALLY required I am not answering a single phone call. Send me a text or drop by in person.
..cant steal a voice if you cant get a connection to begin with
You could just... not answer unknown phone calls. I mean its pretty easy. only answer calls from known numbers and even then, just let it go to voicemail. if its genuinely important the person will call back.
@@shoulderpyro yea forreal. I literally never answer a phone call from a number i dont already have saved in my phone unless i am expecting a call from someone or some organization whether it be a business or other. Its just that simple. I look at people who answer random phone calls with disgust because why the fuck? If its that important theyll leave a voicemail and ill respond when i have the time.
“Rizz finna save humanity, fr fr 😤”
-Kyle Hill, probably.
This comment was AI generated
I hope not!
No, definitely not. He doesn't talk like an asshole
You already know some asshole is gonna make AI that shitpost if not already. This strategy won't work.
@@ldawg7117 How out of touch are you?
As for voice, if you're a video/audio creator, it is advisable to only authorise banks & other businesses to interact with you in person, as noted by that guy that worked at blizzard and is a streamer & shorts producer. Then if they don't follow that notice & do something, they'll be liable because they've not followed the explicit instructions to only accept in-person instructions
PirateSoftware. That guy is awesome and honestly inspiring with his shorts.
@@linkfann17 yeah, that's the guy!
I get the idea behind this, for sure, but whose bank has any goddamn idea what they sound like?? Like if you have a personal banker I guess I get it, but even for an influencer that's weird.
Pretty sure even if you do online interviews and people mess with your voice or video with AI you can still go after them because they are using your likeness without permission.
@@linkfann17next time put an at in front of his screen name. I bet he would get a kick out of seeing the occasional people saying good things about them stuff.
@piratesoftware
I have refrained from staying subscribed to you for the longest time due to your over self promotional and over expectant behaviour (I hate overinflated TH-cam beggars), but I've got to hand it to you... This video is amazing, scary, and shows a much deeper thoughtfulness and depth than I had come to expect of you. I dare say... Keep it up.
Get over yourself
I personally would like to clean out the entire internet. There is too much of everything, and I'm sick of it.
It takes a lot of effort to tailor the internet to you. Most people don't bother, and are susceptible to manipulation.
You poor thing
thats if we go extinct
Why can’t we use AI to clean out the ocean? Same reason why warehouses will be left with junk and the internet will never get scrubbed unless there’s some type of military ‘great reset.’
we should save a copy of the internet, so that we can have a kind of checkpoint to start from if we lose the boss fight
I agree 😂
Too late net neutrality already deleted our archives
Didn't you hear him say how fast it's growing? It's already "too big to save"...
Best comment ever❤
Far too late.
Born too late to invent, too early to explore space. Just in time for the A.I. Chatbot uprising.
😂 bout right
*too. Learn English.
@@NeverUseAnApostrophe sugg digg is that enough enRish for u
Abandon the internet and think of something else.
exploring space will always take generations. Space is big. Signing your great great great great great grandchildren to do it is unethical.
Using slurs will be the only way to tell a bot from human
Already is
More like AYYY shawwttyyy,, u gudd brahhh?? Whatssguudd bud kinda text imo. Its scary how it can be the future lol
On thing thats important to note is that ChatGPT's ability to pass the bar exam and the MCAT does not mean it understands the law or medicine. ChatGPT can pass professional exams because it was essentially trained on the answer key. It learned the statically optimal distribution of words to spit out when given an exam question, but it doesnt actually understand the concepts, and can't apply them logically in novel situations.
When a human passes the bar or MCAT it demonstrates that they understand the underlying concepts and can apply them correctly. When an LLM passes the bar or MCAT it's just a demonstration that you gave it enough practice problems to "memorize" how answers are phrased. It's the equivalent of someone passing a multiple choice test by figuring out the pattern of dots used on the scantron rather than actually reasoning through the questions
Yep. None of those AIs actually have intelligence. They don't think and they don't understand anything. It's why halucinations are impossible to fully get rid of.
You say AI "can't apply them logically in novel situations" yet that is exactly what it does, which is why it can correctly answer questions that were not in its training data. Just like the classic example, an image recognition AI can be trained on 1000 pictures of cats, then given a brand new picture it can decide if it's a cat or not. If that's not applying existing knowledge in a new situation, I don't know what is.
yet...
@@achillesa5894the thing is.. thats literally not how generative AI works. its basically iphone autofill but much more advanced, what ur saying is a drawn out version of "nuh uh"
it's like learning a pattern of colors to us, we can remember the pattern, but that pattern is meaningless to us.
1. Captcha for years wasn't an issue for bots
2. AI can bypass the data cut off date by just browsing the web, or by applying a LoRA
More likely it uses retrieval augmented generation but yes, the cutoff date isn't a concern anymore
LoRAs are good to bring up, most people dont know of them.
you do not even realise that captcha was actually used to train AI... yes ALL this time you had to point out where the traffic light is, you were teaching AI what traffic lights are in different contexts
Captcha was never meant to stop bots, then intention was always to make it more expensive to make bots
@@cyborgchimpy technically that'd depend on the captcha
An A.I. may pass the Bar but it is currently quite far from being good at law. Also, I love being algorithmically incoherent. My wandering and seemingly pointless shopping habits, coupled with using a couple of different languages for searches has really confused my adsense. The other day I got a Spanish language Temu ad for like old lady jewelry. There's a great book called "Feed" which is sci-fi, but it has great tips on being internet random.
Yeah. On one hand it's really intelligent, on the other it has some really obvious foibles. Chat GPT needs a lot of curation to make sensible content from it.
Oh I just picked that book up at the recommendation of one of my old coworkers.
Thanks for the book recommendation. I just ordered it x)
There are cases for which I'd be happy to be represented by an AI rather than the appointed attorney.
what is considered "currently quite far from" in AI terms could be a mere weeks or months away. We need to stop thinking in human terms of references, to instead look at exponentially scalable artificial evolution. What LLMs can do today vs just half a year ago is a wider step than what LLMs half a year ago could do vs when they were first introduced to the public. Think about that, then think about this being a continuously exponential curve - at least until some hard limiter is hit, but there isn't to our current knowledge any signs of such a limit showing up anytime soon.
When I first had access to the social media as a teenager (Like Facebook & Tumblr) I thought it was the coolest thing every but after a week I felt like these media were wrong some how and disconnected myself from them and ended in small gated communities like discord and just mind my own business, these days I'm slow opening my "vault door" to these massive wild lands seeing who and what I find and what I find is pretty wild
"Thou shalt not make a machine in the likeness of a human mind." - Frank Herbert, Dune, about fifty years ago. I wonder what he would think now.
He'd be mortified because that whole prohibition was predicted on avoiding a situation where a small handful of people controlled the masses using AI - and later, from the Butlerian Jihad, the concept of it being tabu to rely on machines to take away your mental faculties, was added to this to create the ultimate sin.
And as usual, we waltzed right into it.
"Guess I was wrong"
He'd be terrified.
Herbert was right. This generative AI nonsense is exactly what he warned about.
He'd probably throw himself in a sandworm.
As an AI language model, I want to thank you for the suggestions on how to pass a reverse Turing test. No one has quite as much rizz as Kyle Hill is an internet personality and science TH-camr best known for his series Half Life Histories, which focuses on radioactive historical events.
I swear people's brains are replaced by chat bot
@@RongDMemer I swear.
There is a massive issue with that in person registration idea, in that it automatically restricts people who for one reason or another cannot access these locations, further preventing internet access in rural or poor areas, as well as countries with extremely restricted internets, like China
If this were national priority scale, a person doing the registering could go to those people where necessary.
They can hire a surrogate to act on their behalf. There will likely be a marketplace of willing humans who put themselves up for hire to do meat space things on behalf of AIs as depicted in the 2013 movie, "Her"
wait till you find out that THAT has always been an issue even long before the internet was even conceived
@@bornachTwo humans being paid to meet in the real world because two AIs had to legally “prove” that they’re human in order to get something done is such a painfully believable dystopia.
Doesn't matter anyways. All that does is open up the space for "humanity mules" people who will go and vouch that they are humans but are really there to sign up on behalf of the fucking robots instead.
This makes me want to unplug my wifi and just start living again
Aside from the scary/depressing subject-matter, I just gotta say, Kyle's 'Discount Thor' look goes great with that kind of 'firelit darkness' aesthetic. Kind of makes you feel like you're sharing a campfire with him in some Norwegian pine-forest, sharing stories and mead as the flickering firelight keeps the wolves at bay...
I got more of a caveman "dark times ahead" vibe. Like if he was deliberately trying to look prehistoric, or "outdated"
Discount Thor 😂🎉
Brilliant
@@MrGibleman, it isn't that deep. All he was thinking about was looking like being in a dark forest as described in the thumbnail
Yeah, the way he portrays himself in the video almost makes it seem like he's in a forest in the dark.
The fun part starts when every person you talk to on Discord can be a potential bot. They can play games with you, quote movies, talk about their lives, hold a video chat, and all this time they never had a tangible physical body. But they certainly have beliefs, beliefs that reinforce your beliefs, beliefs that the bot make see as its own but are in fact the beliefs created by its creators.
Thats messed up to think about.
It's basically the ultimate way to scam people. Take all the effort of getting to know someone and earn their trust and just relinquish it to a hyper realistic AI. Then you have them hook line and sinker for any scam you want to attempt. It's terrifying, plus we already have people mimicking voices with AI to essentially hold fake ransoms and email elderly family members. AI is just going to take all the most insidious parts of humanity and crank it up to 11, because people will have the tools to facilitate it
the bot is trained to output text and not to have internal beliefs
So basically not that different from a real human psychology. So what? Right now people watch that video and also form their beliefs the same way. Does it mean it's 100% true? No. But we still do it.
@@cryrustmusic That's the scary part: sufficiently advanced simulation is indistinguishable from reality, and so at a certain point brains are but a grey shade away.
I retreated to a "digital bunker" years ago.
The comment bots, the ads, the manipulation, the synthetic content, the hatred, the anger engagement. I hated all of it. I left the Dark Forest.
I'm far happier in private servers and actually talking to people in person. It feels much more real, and like a return to almost what the internet used to be like in its earlier days.
I've long advocated for a return to the small to medium communities where we had actual moderation. Then the Kurzgesagt channel made a video about how that probably feels better because that's kind of how early human societies were: just a few hundred people huddled together for mutual support.
If going outside is the best way for humans to "signal", wouldn't the AI's first move be to make going outside undesirable? Either through making the digital space 24/7 bliss, or by making the outside inhospitable?