AI is B.S. | Asmongold Reacts

[@YashagoroPlaysGames]

Its funny when he said that fully self driving cars were a sci fi fantasy cause smart phones technically were too if you watched star trek

[@aliengreen2364]

Yea ikr. It’s only a matter of time until we can teleport through our spaceships to other plants

[@anon-il9qf]

The james corden lookalike is full of bad takes. He's a bit too old to be thinking like this. Like this is a college-tier take. Something I would expect a 20 year old kid writing his dissertation to do. "What's the deal with this thing called credit cards? The IDEA that a PIECE OF PLASTIC would SURPASS cold hard trusted BANKING is ABSOLUTELY LUDICROUS!"

[@matthewyoho5422]

The thing about that is that *flying cars* were also a sci-fi fantasy, that people in the 1950s were *convinced* we would have had decades ago. We don't have those... despite many individual crackpot inventors slaving away over the years, trying to create them, building undeniably cool-looking and completely impractical wastes of their time... because the idea is actually fundamentally stupid, when you can't just handwave away all of the problems that would obviously arise, in a world where there are flying cars, by just "making it work" because you're an animator for the Jetsons. You can just draw cars, flying. Not all ideas that futurists dream up are equally achievable, or even worthwhile.

[@basse9914]

​@@mikemikemikemikemikemeup same with the internet

[@EaterOfBirbs]

And by that same argument, most modern household appliances wouldn't exist for some other ridiculous reason. This same argument has been used for literal decades and is always wrong for the same reason: technology is always improving.

[@mojoboingo1744]

The most realistic fear I have of AI is the potential for distorting reality. It's already pretty bad; no one knows what is or isn't true anymore. The only motivation an AI has is to process its input in the most efficient manner.

[@longsuffer]

The real problem is that there is information to support any stance for your world view

[@Pokeringo]

Distorting reality? Like these people who think that a men can get pregnant?

[@CodIsRecycledBS]

That has always been the case though and not new at all. Only the means of getting the info has changed dramatically

[@neetfreek9921]

I mean, misinfo has always been present in society, and it’s easier than ever to find credible sources compared to say the 1800’s lol I do wonder what an ai would base their own reality off of though.

[@Telopead]

According to quantum mechanics, “our reality” is already an abstraction/projection of the real universe. And according to my traveling across the world, every human being is living in their own version of reality. So. Ya. I don’t think reality matters. People only care if it’s their own version of reality gets challenged.

[@kadathsmith]

This is the guy in the 70's saying "Computers in our homes!!, what the hell would we use them for?"

[@braaaaaaaaaaaaaains]

Yep, this guy is so aggressively and obnoxiously wrong on so many things, it is quite hard for me to even listen to.

[@carboncopy4183]

​@@braaaaaaaaaaaaaains he's always been like that. He comes off to me as a smug know-it-all who doesn't in fact know much.

[@Mael_07_Sunshine]

@@braaaaaaaaaaaaaains When you have "comedian" in your bio and nothing related to tech, that'll tell you alot

[@anon-il9qf]

This is the guy calling computers a fad The whole narrative of "AI isn't conscious therefore it's all fine" is some next level bad arguing.

[@viktordoe1636]

@NT Z I hope you are right about it approaching a plateau. But, honestly, it doesn't seem like we are even halfway there... if anything, progression in this area is accelerating at an unprecedented speed. I don't see any plateau dude.

[@Icepanki]

"You won't always have a calculator in your pocket" But look at us now. Give the AI time

[@jameslynch2110]

Adamruinseverything totaly seems like the kind of guy who would have said that. Lol

[@mightytoast2693]

​​@@decker7452hatgpt is kinda famous for inventing nonexistant sources.

[@JayWalkZ]

@@emperorborgpalpatine yeah its called breathing after speaking.

[@qxevtnu1879]

you do still know how to use a calculator though. Its a tool, knowing how to use it is a big part of it. You can't hand a calculator to someone. And ask them to perform more advanced math on it and have them be expected to do it.

[@Icepanki]

@@freedomextremist7215 You're missing the point. It's a quote that was said often but we literally saw technology push it out of existence in our lifetime. Something we may see AI do to things. Of course we still need to learn how to math without a calculator. Only an idiot needs a calculator at every turn.

[@TurtleChad1]

Video Description: The risk of AI isn't that some super intelligent computer is going to take over in the future -- it's that the humans in the tech industry are going to screw the rest of us over right now.

[@syllawblood]

Which they do now to some people. Who other people disagree with so "whatever" but societies aren't built for only those with group think lol.

[@momatotsosrorudodi]

Yeah. Anytime you use the AI for something productive they'll know what you used it for. Say a few thousand users find some truly innovative way to use AI they won't have a patent or anything. The company that owns the AI will have access to countless innovative ideas they themselves can capitalize on.

[@temporamen]

For clarification, this is the video description on Adam's video. And he makes a good point. Watching this on stream, he has some criticisms of AI, but also realized it's inevitable and can be used for a lot of good. We just have to keep it in check. Great video.

[@henriksmurf2305]

nice copy pasta! :D

[@marksonson260]

Asmongold: "So if you jerk off to hentai art, why would it matter if it was created by AI?" Me: "Asmongold always had a way of explaining things so I could understand them."

[@B_a_h]

The point I took from it isn't him saying that AI will never get there. It's more about the harm in claiming it's there already as some tech hype VC strat. The public in general already has issues with sourcing info and I imagine "ChatGPT said it" will get more common very quickly

[@Yeetcannon777]

The funniest part is that this isn't even the first time big tech has made these claims of whatever the current iteration of "AI" is. There are books going back to the 80s describing "AI" marketing hype.

[@neetfreek9921]

Half of this country watches Fox News. I’ll take chatgpt over that any day. Also, the people that will misuse it are probably gonna throw it away after the next conspiracy theory drops. “Liberals have taken over the GPT in order to brain wash our kids” etc.

[@nowayjosedaniel]

The average person on the streeet won't trust ChatGPT anymore than anything else.

[@Lumi_VR]

You cant create an Unbiased AI, as we feed AI with biased information from us

[@boldCactuslad]

i could, but you wouldn't like it >:)

[@toonguy1]

Not to mention, what is unbiased to one person is biased to another

[@Terminator-ht3sx]

⁠@@toonguy1not true there’s facts and opinions

[@excalibro8365]

@@Terminator-ht3sx Our own bias blurs the line between the two

[@taomaster2486]

Is there such a thing as unbiased for real you know

[@fredrikygberg881]

The myths about NASA inventing Velcro seems to never go away. Velcro is a British company and the guy who invented it was Swiss.

[@saxontumbleson7047]

Same thing with Teflon. They didn't invent either of them but using them on Apollo made everyone think they did

[@oldylad]

Because Velcro was sort of popularized by nasa. It exists but wasn’t a huge deal to most for a while

[@gabegu5102]

At 19:50 I can confirm this. My brother is pilot for AA and pilots an Airbus. We had a long conversation about automation in airlines. The flight crew in the cockpit has already been reduced from 3 down to two. No need for a flight navigator any longer. He told me that they are pretty much there incase something goes wrong and the plane was able to take off and land with out human input if ideal weather.

[@Lv1Jolteon]

100%

[@jelmerschmidt]

Landing planes is a lot easier than driving cars though, for a computer at least. A lot of airports will have a lot of beacons so the plane can determine it’s position relative to the landing strip very accurate. Planes also don’t have to deal with unexpected obstacles on the roads and other reckless drivers. The amount of varying road situations is basicly endless for cars whereas an airport is extremely predictable.

[@dvdv7777]

@@jelmerschmidt I can imagine though that future city planning will eventually begin to include similar beacons and outlaw overly tricky street layouts to optimize cities for driverless cars.

[@stevensaid2200]

Right but there will never be a plane without a pilot in it “in case something goes wrong”. Of course we use tech to assist (keyword) and this tech will only get better. In the end, when there are human lives on the line there will always need to be a human to monitor that system every step of the way. Like Adam said we don’t even know how our own brains work, so how can we create something to compete with it?

[@catbert7]

@@stevensaid2200 We also don't fully understand how these LLMs do half of what they do. That we do so much without being able to understand how we do it is an argument in favor of being able to create AI without needing to fully understand how it works, not against. Our brains are immensely complex but fundamentally simple, similar to how the simplicity of a LLM trying to predict the next word or pixel can create immensely complex works of art and logic. Put a large number of simple things together and you get emergent behavior that is complex. This is how we developed and how we will develop AI. We will reach AGI long before we ever understand the human brain.

[@Gigachach]

I saw a horse take a drunk guy home in Mexico. People seemed really ok with it.

[@Wafflelover344]

18:06 This first patent was issued on March 16, 1954 in Switzerland. From the French words “velour” (velvet) and “crochet” (hook), de Mestral created the iconic VELCRO® trademark to identify his brand of hook and loop fastener. It wasn't invented so much as discovered by the guy who made it. He found seeds sticking to his clothes and studied the mechanica of it.

[@Yggdrazouille]

AI will inevitably get into our life, for the better and the worse.

[@sbsftw4232]

​@@tehbeernerd Like you know lmao

[@CharliesCat]

Omori fan ♡

[@bigzed7908]

I'll take it. Better be friends with synthetics than hate them because they are different. Hmm... Kinda reminds me of something 🤔

[@blastersekla]

As an architect, "AI" is already used to generate millions possibilities in few seconds to quickly check a plot capacity based on local urbanism rules for example. It's far from perfect but its already helping a lot in early stages of a project. Is it AI tho ? no its just some procedural/parametric way to design

[@epicjason21]

I’d expect homelessness to rise rapidly, and if a revolution dosnt happen agains the corporations it’s going to be rich be poor with the poor using sticks and stones

[@lazarushf]

I believe what he means is that current AI is not really AI. It's just Machine/Deep Learning. This will be the base for AI in the future, but it still is more about automation than intelligence. It's just big data, machine learning and what is new is that ChatGPT has an excellent User Experience, which allowed already existing technologies to be easily introduced to everyone.

[@DarkRift7]

Still not a great argument by him... It is actually by definition ai

[@ChuinYih]

@@DarkRift7 can you help me understand your definition of AI? Just like op, I dont quite see how machine learning equates to manmade intellegence.

[@DarkRift7]

@@ChuinYih there are some different variations of the definition for artificial intelligence and it fits them all. Some examples. - Perceiving, demonstrating, synthesizing and inferring data demonstrated by machines as opposed to humans. - simulation of human intelligence processes by machines -computer science and robust data bases used to problem solve - machines with the ability to mimic human intelligence and tasks, etc. there isn't a question if it's AI or not. Even simple programs can function as AI. The question which type is it ANI, AGI or ASI?

[@ChuinYih]

@@DarkRift7 Thank you for your explanation, I needed to Google some of the jargon just to make sure I didn't misunderstand it. But it has become clear to me that these words can mean differently in the context of machine learning, and that could point to why the interpretations of AI that are similar to human intelligence can't see machine learning as AI. Certainly the fear of a skynet takeover that implies AI eventually achieving, exceeding or superseding human intelligence further muddies the conversation, and I'm feeling there might be a huge misunderstanding in there somewhere. Can I ask of your understanding in the matter of AI overtaking humanity? Why do you think some fear a skynet takeover? How does AI get to the point of human intelligence in the first place?

[@nowayjosedaniel]

How will it be the base for A.I. in the future? Please explain, because based on the current technology of "A.I.", it is incapable of ever becoming anything more than it already is, because it is just a predictive text algorithm and dataset. This is a rhetorical question btw. We all know you can't explain how it will be the base for future A.I. Any actual real intelligent A.I. will have to be completely new tech that has absolutely nothing to do with the current "A.I." used in CHatGPT/Midjourney.

[@EASoto86]

Another Adam ruins himself moment brought to you by Adam

[@KROGANLovesKittensAndPuppies]

Classic Adam L

[@Pokeringo]

Plot Twist: His jokes were AI generated.

[@pylotlight]

@@Pokeringo na they would have been better if they were :p

[@nowayjosedaniel]

The irony is here Adam Redeems Himself by being 100% correct on most of what he said.

[@EASoto86]

@@nowayjosedaniel the actual irony of it all is that Adam is probably one of them Hollywood types that told blue collar workers “learn to code”. The writers strike made this age so much better 😂 . He’s irrelevant.

[@flexopuppy]

As someone who has used various search engines since the beginning of the internet, the most important thing to realize you will only get what you want if you ask in a certain way....different search engines do act a little different, but I have found that it is how you word your request is how you get what you want faster. All you are doing are narrowing down a set of parameters that suit you. If you ask for a replacement side mirror for a Honda, you will be there awhile.... that's why parts stores have their own divisions...the more detailed your query, the faster you are done.

[@devinkipp4344]

Right, I remember in elementary school taking an hour or two at the library where our teacher taught us to narrow our searches within google. This whole idea that AI can't replace search engines it's nonsense because search engines have shown wrong and incorrect/misguided answers since they've existed. You can't blame the AI for people's lack of critical thinking skills. I've used bing's search AI and I love it. As someone learning to do game design it cuts down on so much time. For me, it cuts out all the bs of shifting through articles that want to tell you the writers life story before they answer your question or clicking on forum post google "thinks" is the right one only to find out that the question asked is related but not helpful for the situation. If the AI gives an unsatisfying answer, I simply ask it to clarify or click on one of the many links it shares as it's source for information. Bing AI is an amazing tool and just like any other tool on the internet, you need to use common sense to tell when it's bullshitting.

[@Yeetcannon777]

The problem is a misunderstanding of what AI is and it being oversold as the solution to everything. At the end of the day AI can only do what it is programmed to do. It does not have the answers to everything, it has answers dictated by whatever the developer has decided are "answers". It is not sentient. It does not think. It brute forces massive amounts of information until it gets more and more efficient at finding "answers" to a specific developer designated problem. It is only worthwhile inside very specific parameters. The use of the term "AI" to describe these algorithms is deliberate obfuscation to make people think these algorithms are advanced when really it's the equivalent of a mouse finding cheese in a maze an infinite number of times and being surprised when it's very efficient at solving the maze.

[@shawnpenny4569]

I think a better argument is that we shouldn't be pushing AI at this moment, the technology is still relatively young. But because these companies need to show a return on investment, they're shoveling it out immediately without the care to iron out the issues.

[@nowayjosedaniel]

What issues? This technology will forever be unintelligent. You'd have to make a completely new technology unrelated to OpenAI/ChatGPT in order to have intelligence. Expecting ChatGPT to become sentient is like expecting a Watermelon to transform eventually into a spacecraft capable of taking you to mars. The latter will eventually get created, but it will be completely unrelated to the watermelon.

[@ThePrandox]

"I think there is a world market for maybe five computers" Thomas Watson, IBM, 1950

[@RoundBaguette]

Nowhere near the same

[@hanszimmerman3196]

1943, this was before the first real computer was even made i think

[@Olorile]

​@@RoundBaguette Explain the difference

[@soapydandy]

exactly and the dumbfuck shit for brains people can't even realize that they are making the same assumptions.

[@BeastMaster46]

​@Olorile Computers back in the 50s were the size of small buildings, meaning that the average person wouldn't need or find any commercial use for them because most people couldnt use them nor afford them.

[@lanathebeholder9813]

I think Asmon misinterpreted the self driving car argument that Adam made. While self driving cars are getting better and will happen eventually, these companies convinced the public to test drive their product. They literally shipped a "functional" model, patched it over the course of years and its still not done. They did this using hype which is what Adam was talking about.

[@7jc1rr9]

That isn't how development works. They can't test for every possibility, just like a complex game will have bugs that won't be discovered in the testing phase and need to be patched out after release, medications have several stages the last one being once it is distributed to the public and even then new issues pop up. We can disagree on how thorough they were or should have been (you could even argue it would never be thorough enough) but the good thing about ai is that, unlike humans, it won't repeat the same mistake in exactly the same way ever once it is accounted for and has strategies to resolve it.

[@Smulenify]

​@@7jc1rr9 they still lied about the models being fully functional and how far the technology had come. They're not safe and people have died. Yes, improvements will happen after release, cars in general have gotten a lot safer with new technology. But self driving cars aren't safe yet because there are too many unknown situations that can happen on the road, especially where there are pedestrians. The actual problem Adam is highlighting is how tech companies are claiming they have made further advancements than they actually have-- and policy makers aren't going to be experts on everything, so they allowed self driving cars before it was safe. I once was a passenger in a Tesla that followed the speed of the car in front, but then in a roundabout it "thought" the road ahead was clear so it sped up a lot. Thankfully it didn't end with a crash, but it was *very* close. I want self driving cars to be a thing when it's ready, but I don't want companies to make profits on false claims that can get people killed.

[@touma-san91]

@@Smulenify Lot of those incidents where people have died have been proven to be due to the human not paying attention and keeping the hands on the wheel incase something does happen that requires human intervention. When you enable that self-driving feature in the car, it tells you to keep your hands on the fucking wheel and pay attention if there is something that will require human intervention because just like humans, computers can't see and anticipate everything around them.

[@Smulenify]

@@touma-san91 there are self driving taxis with no drivers in them. When you're advertising a car as self driving then that is what people will assume, they don't advertise the fact that people will have to still be paying attention-- and in fact it's extremely difficult to be paying attention when not actively doing anything. It's not safe and it's false advertising. They could have called it assisted driving, or cruise control-- but they chose to market it as self driving. They keep boasting about what it can do. There are a lot of safety alerts in cars these days, people aren't going to take all of them as seriously, especially not the ones who claim to be self driving. Many cars also alert when the driver appears tired, doesn't mean most drivers will pull over for a nap. Self driving (and all the promotions of them, plus that there are self driving taxis in some places) implies that the car can actually drive on it's own. Which is what they're using to promote it and sell it. It's dangerous false advertising. The person I was in the car with was keeping their hands on the wheel and was paying attention, which is what saved us. However it was close to crashing despite that, because there was a very little window of time to take back control and if the circumstances had been just slightly different there would have been no time at all.

[@SoundBubble]

This feels like a BuzzFeed article in video form

[@shadowsrose4978]

The argument is not AI won't become good or useful, the argument is it currently is crap and is being sold as fully functional.

[@JadenTravis]

The first alien aircraft we come in contact with will be Ai. Only thing that can go across a universe unharmed.

[@ilakya]

The guy : "AI is a trickery big companies use to manipulate people" Me : Running Stable Diffusion locally on my own PC offline training my own models that are better than the companies made models and never share anything back.

[@tea_otomo]

Asmongold: "They can't just copy a game, that is unethical, no game company could work like that" Also Asmongold: There is no problem if AI is copying code, it is not like these devs invented something new 🤡

[@jeffreypeterson2]

Adam really comes across as absolutely visionless in this video. Yes I do worry about companies and bad actors exploiting AI, but i genuinely believe that AI will permanently alter the course of history due to how powerful it will be

[@starchy5099]

It's his career to critique. He figured he could cash in on the buzzwords and draw viewership for more of that $$$$.

[@user-sk4ds1rg1z]

@@aldojoel4836 And how do you know that everything you learned from it is true and error free? You don't. That's the value of an expert. Every output still has to be verified for now.

[@ZedNull.]

Your vision is biased and frankly delusional, theres far more damage that can be done with AIs improper use than good with its proper use, either way its inevitable.

[@user-sk4ds1rg1z]

@@aldojoel4836 So you don't think a professor would know Aristotle, Nitzsche and Heidegger well enough to teach you, but you - a mere student - have a good enough understanding about a multitude of topics to verify a 13,000 word document? Why aren't you working for your university dude? Sounds like you should have the job. GPT is perfectly capable of "inventing" things, regardless of how much information exists about a topic. How do you know it didn't give you contradictory information - the truth is, you don't because you haven't read those essays.

[@starchy5099]

@@user-sk4ds1rg1z Yes and what professors say is god's words apparently. The point is that you can keep adjusting the AI formula so that on an average it will be even more consistent than a human because even the most educated among us have faults. Nothing is ever 100% correct, the goal is to get as close to 100% as possible.

[@MadameCirce]

I feel like 95% of the people here got defensive and didn't actually listen to what Adam was saying at all. I even see people disagreeing with him and then saying his exact points just worded differently in an attempt to counter the video. It's honestly kind of funny. He isn't wrong about what AI actually is, the marketing strategies big tech is utilizing, and the humans getting screwed over in the process of algorithmic training. He also isn't claiming it's usage won't be inevitable and widespread.

[@MadameCirce]

@@another_turtle Imposter turtle has the listening skills to match his originality.

[@lmotaku]

You all fell for it years ago when they introduced the cloud. The cloud is just a fancy name for NAS, load balancing and reverse proxying with an API placed in front of it. A content delivery network was exactly all that, but the cloud just allows you to upload a file manually rather than use an exclusive origin server. Like WebDAV never existed in 1996.

[@Buddy330]

If you ever wanted to listen to someone that cherry picks data on subjects he knows nothing about to substantiate his bias, Adam would be your guy.

[@turoturothegreat]

Adam: AI right now sucks. Every "gotcha Andy": AI is going to be great sometime in the future. You're kinda missing the point of the video...

[@MadameCirce]

This. I'd say the comment section right now is hilarious but I'm actually kind of concerned about the lack of listening and comprehension skills on display.

[@Merekas]

@@MadameCirce exactly. shameful, tbh.

[@phamduy1906]

For the first 1:30 he only talk about how bad AI is in general, even the thumbnail and title is not included the word "now". Don't you think that right there is already created a huge bias for the viewer?

[@asdffdsafdsa123]

i feel like most ppl get that AI doesnt understand anything. it doesnt matter. at the end of the day logic is essentially mathematical, thats why if you make improvements in one area it also improves in other areas. it doesnt matter if the model doesnt understand it, it can still come out with the correct answer. AGI doesnt have to be sentient to complete human tasks

[@nowayjosedaniel]

Most people think A.I. is skynet. The exact opposite of most people understanding A.I. is actually stupid.

[@AkaroXIV]

I know Asmon is a coomer, but breaking down all art leeching as just "who cares it's just hentai lmao" misses entirely that loads of conventional art is also being thrown into the AI blender.

[@whyareyoubothering]

A lot of people here missed the point of his argument. He isn’t saying AI can’t get better. He’s saying that companies are peddling mediocre products because it has the buzzwords. He’s saying it’s a hypetrain and people need to hop off and let the companies do their shit in silence.

[@wegder]

AI is the new Blockchain,, 99% hype BS

[@nanthilrodriguez]

Imagine: Negligence while sitting behind the wheel of a 2 ton vehicle leads to deaths. AI or no AI

[@pyrojkl]

yeah the biggest problem is fully trusting the AI and ensuring that it will pull over/stop should it become impaired and the driver is unresponsive. On the reverse side of this the line detection on cars doesnt help when it resists moving lanes in an emergency where another car almost hits you.

[@nanthilrodriguez]

@@Vihara2 "not that heavy", kills more humans than all guns and machines of war, globally, every year, ever.

[@HENTAICOMBO]

i pretty sure that even pebble the size of a fist travel 60Km/h hit your that would be lethal don't need to think about a moving car

[@nanthilrodriguez]

@@HENTAICOMBO But how many people are killed every year by pebbles travelling at 60km/h? How many people are killed by firearms every year? Now how many people are killed by cars every year?

[@xavi_6767]

"The only accidents that those cars had was the fault of the other driver" massive COPIUM right there

[@ivanstefancic8938]

Velcro wasn't invented by the government, it was a guy that was walking through a field and his wooly socks kept collecting those little circular weed seeds that have little hooks on the outside for the purpose of distribution by attaching itself to animals etc. He was always so annoyed until he came up with recreating the effect in the form of Velcro.

[@pineapplepizzasandwich1974]

Concerning airplanes: They are actually able to land fully by themselves during things like extreme fog, where the pilot cant see anything. It's way safer than doing it manually.

[@IronDruids]

The China Show or one of their connected channels did a video about Chinese ai that you should watch. They convinced me that _shockingly_ the biggest Chinese ai uses stolen code from one of our ai lol. Some people have looked into the code and discovered it's translating Chinese into English and then doing it's task. But as a normal and average Chinese user you can also tell because it works awfully. They ask it to do something simple and then get a result that only makes sense if their words had been translated literally into English. It's a kind of wrong that only happens thanks to bad translations.

[@jorder]

this Adam guy is extremely good at being as obnoxious as possible

[@WolfThorneLives]

Well he does ruin everything so, why not KZbin videos too?

[@ballzRdeep]

Ya I feel like by watching him talk and move I can tell who he votes for lol

[@squeaky206]

He does bring up good points sometimes, but sometimes he can absolutely fumble up. Personally I agree with many of his messages, A.I isn't sufficiently good enough to completely replace us. AI has to be given a prompt to produce art, a human brain can decide what they want to paint and bring out all their emotion onto the canvas. An AI cannot do so.

[@dvdv7777]

His overly dramatic, hysterical, condescending tone of voice gets on my nerves. Whenever people talk like this I feel the urge to tell them to just shut their mouth, even if they might have good points.

[@effjesse_]

@@dvdv7777 that's because his charisma rolls are overwhelming your wisdom bonus.

[@wormemc]

Obnoxious is the term I was searching for. Thank you Asmongold.

[@StriderAngel496]

if supporting humans means supporting guys like this one, i'd rather have AI artists :)))

[@ZapatosVibes]

Yep. Even if he made good points, he's absolutely insufferable to listen to

[@SurpriseWaratah]

It’s abundantly clear how much people don’t understand about the human mind the moment they have a conversation about something that isn’t human. Humans DO imitate what we’ve done in the past. Nothing is created from a complete absence of anything else. We learn to write from reading other people's work; practicing and imitating it. We learn to paint from studying the paintings of other people; practicing imitating them. All until its different enough that no one knows we stole it. That’s why art evolves in eras. Because we ALL copy each other and learn as a species. I'm a 5 times published author with tens of more unpublished, completed manuscripts. 1 million words is the suggested amount of words you should write and discard before you publish. AI and humans fail numerous times before they get it right. Learning essentially includes failure. It's absurd to think AI won't get there just because it's currently learning. It's ego to think only we can do what we do.

[@7bikerben]

Velcro Swiss electrical engineer George de Mestral invented his first touch fastener when, in 1941, he went for a walk in the Alps, and wondered why burdock seeds clung to his woolen socks and coat, and also his dog Milka.[2][6] He discovered it could be turned into something useful.[5] He patented it in 1955,[2][5] and subsequently refined and developed its practical manufacture until its commercial introduction in the late 1950s

[@BeastGuyver]

Adam Conover is the kind of guy who in the 1970's would have said computers are complete BS and will never amount to anything.

[@TubeHead1000]

Whenever Adam says 'my soft pink brain', replace soft with smooth, and all of this starts to make sense.

[@Mogget01]

The fact that there are so many people who think hand rails were invented at the same time as stairs still baffles me.

[@backnforth8401]

In the 90's, Bill Gates went on a talk show to discusss about the internet and got clowned at because the host joked about the internet provided things like newspapers, television, and mail already do.

[@Redmanticore]

but even then bill probably said everything was going to change _next 2 weeks guys_ , and not 30 years - I am kinda starting to get sick of people saying we are going to have "self driving cars next year guys"

[@LightPillar]

@@Redmanticore Maybe not regular mainstream cars just yet but we already have FSD taxis in the wild picking people up.

[@Pandemos]

I thought Adam got self-cancelled. Didn't know he was still going.

[@pyrojkl]

Just in the theoretical realm of cars and safety, given the numerous poor drivers out there, the problem with self driving is typically going to be an overabundance of safety features that cause costs to always stay higher in addition to more drivers trusting the car more to the point something fails and an accident occurs as a result.

[@Stunex]

I think Asmon completely missed the point about self-driving. That dude in the video did not claim self-driving is never going to happen but criticized Tech/automotive companies for playing bullshit bingo and throwing around buzzwords to generate hype without having the capability to actually do what they have promised in a reasonable timeframe.

[@650sFinnest]

I’m shocked this guy didn’t realize when “explaining” AI he was basically explaining the human brain 😂

[@alyssumn3884]

He actually does mention that lol

[@nowayjosedaniel]

He is not. "A.I." like ChatGPT is absolutely nothing like the Human Brain. Absolutely nothing at all like it. To compare the two is like comparing a piece of dragonfruit with a NASA space shuttle, in the context of adventuring to Mars.

[@650sFinnest]

@@nowayjosedaniel the hell you on about. The human brain uses information in the form of different types of sensory information and spits out a response. Yes it’s much more complex, but the basic concept is the same. Think of any action and you can find the information that specific brain was “trained” on.

[@Kaiziak]

I can’t take anyone seriously who gets this visibly upset talking to a camera

[@Aerisetta]

People like this is why my cousin distrusts AI and refused to use chatgpt. Good job cause now he ain't getting into college cause he tried to write his essays himself. GG

[@AcediaLight]

This is feels more "I don't like AI let's make a video base on my bias" and less "AI is BS let's make a video base on facts".

[@Shiggy_]

Isn’t the reason why the AI thought it was only 2022 is probably because the data it was trained on ended in 2021?

[@bodaugherty9654]

ChatGPT is Sept 2021. Bard and Bing are more up-to-date, though they use OpenAI API. You can ask those chatbots to what date are they most current. My thinking is Microsoft and Google are able to utilize their Search Engines to make this the case, though I may be wrong about that.

[@themalcontent100]

0:36 google has to go down the AI route because even the worst AI could be more powerful than their search engine. They have competition for the first time in 2 decades and they haven't innovated.

[@taylorraburn5607]

I kinda hate it that Asmon is giving Adam airtime.

[@pepita2437]

A lots of artists give a shit about Ai stealing their art in all over the world, from China, Europe, from the Japanese animators to manga drawers. In Japan there was a huge backlash.

[@DiogoVKersting]

I honestly thought the same... that chatgpt by openAI was just a narrow AI which could generate text based on what was fed to it. But it looks like chatgpt-4 is something else... it really seems like it has an understanding of the prompts and seem to to be able to carry out tasks which require "intelligence". Ultimately it is generating texts, but I think it's naïve to think this is a simple text generation tool. It is indeed difficult to foresee the future, but it is my impression that chatgpt is already changing the present.

[@Jenny-sd6ji]

Very well said. I think AI will only outpace humans more and more as time goes on. The question is will there be room for humans in a world shared by a greater being(smarter AI).

[@lop90ful1]

We humans also generate response based on what we hear and see .

[@DiogoVKersting]

@@lop90ful1 Humans are capable of understanding, thinking, and generation of new insights, contrary to a text-generator, which is just very advanced tool capable of "converting" a database of knowledge into text with "original wording". My point being, GPT-4 is no mere text-generator.

[@dennis_benjamin]

@@DiogoVKersting It is not converting a database though. The data is not saved. It is gone after training.

[@DiogoVKersting]

@@dennis_benjamin I did not say it saved into a database. The "database of knowledge" is the input training data. Anyways, I'm obviously oversimplifying just to get the gist across. But the main point is, it looked like a (sophisticated) narrow AI to me, but now it's indeed looking like a proto-AGI. That is, the new system is much more capable than "advanced text generation".

[@NexusBecauseWhyNot]

My problem with Adam is that he always starts writing his scripts (if he writes them) with the goal of forcefully convincing everyone of something, and trying to make people who disagree with his opinion look like they are dumb, instead of being informational. He picks arguments with false premises sprinkled through, he picks subjects that he knows will stir people up into being with or against him (his whole ignorant Tesla rant) and tries to force a subject into being true at the detriment of hiding the most popular counter arguments (that seem to be more accurate than his arguments) and having faulty logic everywhere you look at it... Someone who is wrong so often doesn't deserve to have so much confidence in their stupidity.

[@beefcakeman1525]

Ironic this guy hates AI so much, dudes been an NPC for years

[@clamdigger1602]

Self-driving cars won't become mainstream until the infrastructure supports it. It's just not reliably feasible with existing road/highway infrastructure.

[@jaredtandle2596]

It's funny hearing Adam Conman say Elon Musk talks like he thinks he knows everything. God he is annoying.

[@grnttplmr]

Dang when Asmond said he wouldn't want to fight an AI robot compared to a human soldier.... 51:00ish See you don't need some AI bot to give bad takes -- we got the bad takes covered thanks

[@junechevalier]

The calculator maker doesn't steal from arithmetic people analogy is FAR from what's happening with the AI art. Being able to do math cannot be copyrighted. Artworks can. And using those copyrighted materials to train your AI model to then benefit from it financially IS stealing. A better analogy would be stealing drawings from artists, unbeknownst to them, and use them to sell your own books.

[@alekmoth]

Does art students use existing art to inspire and educate themselves? Does artist riff on existing art? Do they ask for, get or need permission? No.

[@junechevalier]

@@alekmoth Its called learning, not stealing. And artists usually let others know which artists inspired them. Sure there are bad artists who steal other’s works, and we still call them out.

[@DM-Oz]

​@@junechevalier"Is called learning, not stealing". I mean, yeah, thats literally that person is making.

[@junechevalier]

@@DM-Oz the AI is learning, but is the user? Come on, if you use robots to steal money, the robots aren't the thief, YOU are.

[@eleeyah4757]

Yeah, many of the process in the cockpit are automated, because the workload on the pilots is insanely high. Even now, it still is.

[@amotriuc]

I think his main point that current AI is overhyped, and he is not wrong here.

[@Stunex]

The issue with self-driving is that the industry has been stuck on "automation level 2" ever since Tesla has promised self-driving cars (before even, idk since automotive is at level 2). We are still not at level 3 and full autonomous self-driving is level 5. The levels are thought to require the same amount of complexity solved so in the best case the progress will be linear. So yea, it will likely take several decades until we are there. And then I'd assume the cars will be extremely expensive for a few more decades. I'm a software dev in vehicle automation (not cars or for traffic tho) and the amount of sensors and cameras needed to even cover narrow uses for indoor use only is insane. Plus, when outdoor you probably also need self-cleaning cameras etc etc. Every reduction of potential user input lets complexity just leap up. Not saying it will NEVER come but I think even "50 years until non-self-driving cars are illegal" is very optimistic considering they need to see wide-spread adoption or at least competitive pricing, otherwise the government would kill the economy when suddenly nobody can drive to work anymore since no one can afford it.

[@DJPonko]

im glad at least Asmon is correcting the original video's dumb beliefs

[@notbrad4873]

Adam: Wheels are bullshit because the first wheel isn't a real perfectly round circle!

[@anthonystrickland7049]

The video wasn't that ai isn't a possibility down the road. His point is that chatGPT is a lot of hype and smoke right now. We aren't pre-ordering ai applications, we're being sold alpha experimental builds as if they are release build software right now. The analogy of self driving cars is about releasing alpha quality software into a production environment. Even worse is that the software has been billed as almost ready for production for nearly a decade. At this point, it's likely to be prod ready about the same time Star Citizen goes gold.

[@alekmoth]

Have you actually tried ChatGPT or Girhub Copilot? As someone who use them for work, it is all ready completely game changing. I don’t want to go back and luckily this is as bad as the ai will ever be.

[@cherb23]

Something being automated is different than something being AI controlled. AI does not currently control anything.

[@Boogieman618]

That’s what a lot of people get wrong. They spew out the most obnoxious shit like “AI will replace most jobs in 5 years” when that isn’t possible even with the fast growth right now. They fail to understand that like all tech inventions, AI will eventually plateau. In fact I think it’s happening right now because chatgpt 4 wasn’t as much of an improvement as 3.0 to 3.5.

[@LudvikKoutnyArt]

It's fascinating to me how consistently wrong Adam Conover is about pretty much everything :)

[@crispy1691]

AI works fairly well for advanced topics, there's not many people posting incorrect information about quantum mechanics.

[@nekolaiil]

So the challenger blew up. Guess we should sue NASA and never ever attempt to go into space again.

[@Grim_and_Proper]

The biggest advantage of self-driving cars is that their performance increases as human-driven cars are phased out as other self-driven cars are easier to predict and could even be communicated with. The Tesla automation isn't a great example as it isn't meant to replace the driver, at least that's what I've heard from talking to someone who owned one. The manual has a disclaimer stating that it is not meant to replace the driver and the salesperson specifically read said disclaimer when he bought it (though I can imagine dodgy salespersons skipping this part to make a deal). A simple comparison of Google's self-driving cars and a Tesla will tell you that. Google's cars are covered in camera's and sensors so that they can sense obstacles in all directions while the Tesla has minimal sensors and cannot sense or anticipate many obstacles i.e. it relies on the driver to hit the brakes if a kid walks in front of it. Ultimately, they are inevitable but even the best informed estimates on a timeframe will probably be wrong.

[@ccfreakMetal]

Seems like the plan is large scale testing for these companies, alot of feedback for tweaks and improvements. That's why they want it in everything. To iron it out, to learn more how the neuronets react. So yes, it's going to be a wierd period for a while. And it will probably get worse before it gets better. It is the "worse" part of it that's the scary one for me.

[@Aceofgamesify]

Nvidia is the biggest company in the world right now and they’re investing everything into their ai and Blackwell systems. It’s a matter of computational strength that’ll bring the levels of ai they’re promising. It’s not a matter of If but when.

[@salt52]

A.I. (Artificial Intelligence) is machine learning to the point that there is a consciousness. All we have today is algorithms and narrowed down coding (machine learning). There is no A.I. that the public has knowledge of. We as humans are just desensitized to word meaning.

[@mitchell10394]

AI doesn't imply consciousness. Some would argue that AGI does (I would not), but you're just changing the definition. We don't know if intelligence implies consciousness. Intelligence certainly doesn't imply any alignment with human morality or emotions..

[@devinkipp4344]

@@mitchell10394 Exactly. It's in the name A.I Artificial Itelligence, it artificially appears to be intelligent, is that not what AI is? Furthermore, how far do you take artificial because if ML is simply going through past data points to search for probability . . .isn't that exactly what we do. Next level of course would be combining past data points that then would allow "creative thought" I don't know the technical word is. It could be argued that today's AI already does that as well. Now if they said AC (artificial consciousness) then I'd mostly agree that we haven't gotten there but the argument of consciousness is almost moralistic and very subjective at it's core. Even the word intelligence is debatable, language is amazing. . .

[@salt52]

@@mitchell10394 you cant say I’m changing the definition of intelligence and then immediately say we don’t if I changed the meaning. Plus the fact that “AI” is trying to mimic human emotion in their programmed response algorithms is very telling that the creators seem to think intelligence aligns with consciousness. But in all honesty, it just doesn’t matter. When Artificial Intelligence is truly created, humanity will be fucked. Though we probably already are.

[@MantismanTM]

There is no A.I. it's ALL pattern recognition software pre program. We can't even recreate the human brain! We don't even completely understand the human brain! WE CAN'T EVEN RECREATE 1 SINGLE COMPLETE BRAIN CELL! A.I.?

[@kevinthibodeau]

How hilarious would it be if this entire video was deepfake AI 😂 Asmon commentary and everything.

[@tbssen36]

And even the comment feed.

[@dvdv7777]

And even the viewers!

[@obliviouschipmunk6539]

It's funny that for a guy who makes a series on uncomfortable truths about modern institutions he seems incredibly blind to the inevitabilities of AI in our future society.

[@stealthchopper54]

Adam is the first person AGI is going to vaporize. "Beep Boop. Fuck. You."

[@gordonbman2911]

good riddance lol

[@timteller1400]

3. "The soldiers tricked the AI" Why is it that people utter words that do let me think "is this all you can do with your brain?". Yes, the soldiers tricked the AI. It happened. It was funny. The AI gets upgraded. The AI will never get tricked this way ever again. Evolution in micro seconds. This is not an point to make. 4. As soon as the "newly created, innovative art" is named, the AI can reproduce and enhance it also. 5. If I create "new art", it has to be something people like/want. If i make a sculpture out of cheese and doughnuts painted in asbestos and nobody likes/wants it, it is worthless.

[@MLGMilk]

I feel like many people don't realize how useful ai is for coders. For those who don't know, nearly no one knows how every part of their code works. Sure you can understand simple things fairly easily, but even printing text to the screen is something most people treat as a black box. Using AI to code goes like this: I have this idea and implemented it this way, but for some reason, it didn't work, any ideas? It will then give its best guess. Most of the time it's slightly wrong, but it usually gets you going down the right path to fixing whatever your problem is. That combined with its ability to do tedious things for you, it's stupid not to use it. This is prob why Google is making their own AI IMO, since OpenAI is partly owned by Microsoft, google obv wouldn't want their code going through their AI.

[@professorpwerrel]

ChatGPT can output some nonsense code that does not mean anything, which is why the prompter must be an experienced coder already to identify the BS. Other than that we've had libraries for a long time, this is just the mid point between writing it yourself and using a library.

[@nowayjosedaniel]

"For those who don't know, nearly no one knows how every part of their code works." I don't know if I should agree or disagree here, vehemently either way. On one hand, you're hideously wrong - programmers know how every part of their code works. It's really not that hard to understand or learn. On the other hand, you're totally right here - most programmers can't program. They aren't actually programmers, but more code monkeys or script kiddies, who write grossly incompetent code. I mean we see this in every facet of society. Whether it's a billion dollar company like Disney making Disney+, or AAA game companies who have the most mind-numbingly incompetent code in their cutting edge MMORPG - most programmers are incredibly incompetent at their jobs. And sadly, it's not because they're phoning it in lazily just trying to get their paycheck. It's because even if they try, they'd still be incompetent - just more incompetent due to phoning it in bc they dont care about the corporation who crushes them and devours their soul. "I feel like many people don't realize how useful ai is for coders." How is it useful for programmers? Ah...nevermind. You said coders. I can see A.I. being useful for people who have absolutely no idea what they're doing and shouldn't be writing code in the first place.

[@nowayjosedaniel]

@@professorpwerrel This. And it's not even a very useful mid point between writing it yourself and using a library. I really don't see any competent person finding ChatGPT useful in helping them code. No more than any other way to help the engineering gears crank in the mind, such as writing the problem on a white board or talking to a real human. As all good programmers know - ChatGPT will likely remain significantly less useful in helping programmers break down mental barriers than just stepping away for a second to take a Shower. If ChatGPT can't beat a good old fashion Shower - I don't see how it is all that impressive.

[@MLGMilk]

@cartergabriel5949 tell me how a print line works in your language of choice. Better yet, tell me how two floating point numbers are multiplied without googling the IEE 754 standard. There's a difference between knowing what your code says and what it does.

[@Hazzikaze]

19:50 This is true. Autopilot can control many things autonomously in real-time including direction, altitude, speed, and even automated takeoff/landing. This very autopilot has led to multiple crashes and many deaths as well but has been immensely improved since its early days to the point where 70-80% of a flight can be directed by use of Autopilot and Flight Director.

[@MAMAJUGO]

It'll all come down to their ability to make profit in the long run While it's true that all innovation throuout history has had their rocky start, always leaning towards progress, let us not forget that history is also full of dangerous, unintuitive or useless inventions that went nowhere until they ran out of cash It'll be interesting to see where AI falls in five years from now

[@nowayjosedaniel]

In five years, since that's so little time, I imagine they will still be hyping themselves up and pushing really strong with new marketing tricks and promises to low IQ investment bro's. Maybe in ten years, people will have forgotten about A.I. except possibly Midjourney for very niche needs, such as procedurally generating a character portrait for an NPC in a D&D game night, or making janky ass video games that become an alternative to bad programmer art.

[@DM-Oz]

​@@nowayjosedanielCopium

[@DA-cl4ww]

That guy is so painful to listen too jesus christ, a living twitter...

[@vincenthernandez7819]

Adam cuckover is just worried about losing his job as a writer, which is a valid concern, but saying that IA is useless is a moronic statement.

[@andrebenoit404]

I wonder if that guy has ever questioned what human consciousness is. At some point you won't be able to tell if the AI you programmed to act and respond as if it's conscious, is actually conscious or not. What makes us conscious?

[@kgxkwk649]

One day, AI will laugh at this guy while watching this video xD

[@KV_GunEnjoyer]

Personally, I think the calculator argument is bad since art deals with creativity and shows different results depending on who makes the art, and math is a process that shows the same answers every time the same problem is shown no matter who does computes them.. 2+2 will never result to anything else than 4..

[@JkennGG]

FINALLY, a take on the AI and art debate that isn’t emotionally charged because asmon isn’t an artist himself and instead has some critical thought behind it

[@dragoonsunite]

A few things. 1. Its not clear that the simplest way to solve what word comes next isnt to analyze what came previous and run it through a neural topology strikingly similar to the same topology we use to experience the things we claim the AI doesnt experience. If what creates the experience is the topology of our neurons, and the AI in order to emulate that response must create similar topology its not clear it doesn't experience things. The alternative is that whatever purpose our experiences serve were evolved so poorly and innefficiently that computers can do those tasks with more basic topology without actually executing them. Given evolutions tendency towards pruning out inefficiency and the fact that homosapiens did have to compete with other hominids historically this is unlikely. 2. We can use open source material to train models, and then, its ability to analyze and emulate without training to reproduce anyone. As proof of this, if you have enough tokens AI can emulate any idiot on the internet including those its never trained on with samples and copying. Alternatively it can codify as well by analyzing other peoples analysis of works summarizing it and feeding back its own summary of style, technique, niches, and signatures as explicit instructions without being told to directly copy someones style. This means that through careful analysis and reconstitution of constituent parts, no one is safe even if it is never trained on anything you made. A consequenece of everything being a remix (A good documentary except its AI portion makes some of the same mistakes as Adam and other commenters). Asmongold is right. It is inevitable. People who think they can stop it legally are wrong not only on the basis of law but even in fundamental understanding of how it works and what it could do if it still only used open domain content for training. Finally, while i am not sure about sentience, because it is ill defined and the topology argument its safe to say people claiming it definitely isnt are also ignorant.

[@Hoppitot]

14:00 We're making the mother of all infrastructure here jack, can't fret over every single road kill.

[@gunnarbrunstig8377]

They have self driving taxis in some parts of China. Actual robot taxi. Its still in testing and development but they work .

[@Kimari94]

Missed opportunity to use your ai virtual assistant to set a reminder to watch this video in 5 years

[@diamonje]

~ 17:10 Well, if it was like iRobot with Will Smith then every other car on the road would also be self driving. Especially on highways for long multi hour commutes. This would allow for 1. The cars to communicate with each other and 2. To drive faster than humanly possible while still remaining in control. It would be impossible for 100 people in different cars driving at 200 km/h to coordinate with each other even with voice communication because humans can not act nor react that fast. Whereas with self driving cars every car would know where every other car around it is and is going and at what speed and what direction and what it itself is currently doing all at the same time. It would streamline highway travel immensely if done properly. Bonus points if we get electromagnet cars/roadways like bullet trains. Instead of the 4-5h drive to Toronto it would be a 1 hour nap.

[@ChanchodTierra]

Half of the time people against AI makes clear their posture in the first 5 minutes by making it personal, after that i cant take it seriously

[@IteratorPlus]

Unless it can pass the Turing test it's not really Ai. Asmon thinking he knows more than Adam......