Transcript
Title text: This is how you all fucking sound
[A smug tech bro wearing a sideways cap, watch, chain around his neck stands in front of a data center by a lake with dead fish. A smoke stack blows pollution into the air]
Tech bro: AI is already here, there’s no going back.
[A smug man in a suit with cigarette in hand stands in a restaurant while two disgruntled diners cough from the smoke]
Suit: Smoking indoors is already here, there’s no going back.
[A smug man in a top hat and suit stands in a factory with two sad and dirty children]
Hat: Child labor is already here, there’s no going back.
[A smug plantation owner stands in front of a field with with two angry slaves]
Plantation owner: The Atlantic Slave trade is already here, there’s no going back.
Thats not how it works.
A better example would be “nuclear arms are already here, theres no going back”
Its not a capitalism thing, its an arms race thing.
Once one country starts making nukes you cant stop everyone from following suit to protect themselves.
Same goes for AI, once one country starts doing it, everyone else is gonna need to keep up so they dont lose the arms race.
Yes, but at least at the end of the day you can use nukes to blow stuff up. Presumably your enemies.
If your enemies win the generative AI “arms” race they can use it to, uh…
???
(Yes, I am aware there are military/governmental applications for neural net learning technologies but they’re the types of pattern recognition and signals analysis stuff we already do without needing to build a football stadium sized datacenter every 50 miles and burn the entire nation’s GDP on electricity generation. Most of the other applications appear to revolve around a regime using it solely to shoot themselves in the foot, e.g. powering a fantasy army of likely to be highly defective murder robots or using it to propagandize at and spy upon their own population in order to ensure a ready supply of destabilizing internal dissent always exists.)
LLMs are not the final state of AI
But LLMs are not the path to the final state of AI, either. And that’s assuming only if — and this is a very big “if” — a true general artificial intelligence can even be created using traditional silicon computing methods in the first place. Blithely assuming that it can be is really rather asking past the sale.
Yep, by design LLM cannot become ‘inteligent’, you can only make it more believable but it’s still copying humans not really thinking by itself. No amount of development or money invested will change that, it’s not a pokemon it won’t just evolve into something different one day.
And it’s worth reiterating, the current crop of generative “AI” is incapable of producing anything new or novel. All it can do is reassemble existing strings, tokens, and patterns in slightly different ways. Innovation can never come from such a machine. That will have to come from a human.
The current push is the notion that “hyperscaling,” i.e. throwing even more hardware and space and power and money at the same concept, will magically make it something it isn’t. Obviously that’s not going to work. It’ll allow grifters to make a ton of money over it, though!
Depends on your definition of novel.
From TFA:
The AI did not prove that its approach is the best anyone can do, though. In fact, mathematician Will Sawin has already improved upon the AI’s grid.
OpenAI privately contacted Litt, Sawin, Gowers and a number of other mathematicians to verify the LLM’s proof. Together (and without the company’s direct involvement), they wrote up their individual takeaways. (No external experts have seen the AI’s original output, however—just an edited version of its train of thought.)
What stood out, they said, was the AI’s preternatural patience and focus.
…
“AIs have an edge: It’s not just that they can try all known methods,” says Jacob Tsimerman, a mathematician at the University of Toronto, who was not involved in the work but was part of the companion paper solicited by OpenAI. “They can play for longer and in more treacherous waters than mathematicians without getting overwhelmed.”
…
The mathematical tools the AI used here are not novel, although their application in this domain appears to be. “The model did not invent something fundamentally new that nobody saw coming,” says Sébastien Bubeck, a mathematician leading OpenAI’s mathematical explorations. “It just executed like an amazing mathematician.”
So, it’s a monkeys-on-typewriters situation with the computer able to try and reject the hammering of who knows how many square pegs into round holes until it finally arrives at a workable conclusion, which a human has already bested. And we’re not allowed to see its homework.
This is categorically failing to set the world on fire, except possibly in the literal sense.
To strongman your argument, “LLMs with a supervised training pipeline cannot become intelligent”.
RL training pipelines are much more open-ended, and experts still unsure one way or the other if an LLM + RL could lead to intelligence.
Not only that, it’s basically eating all the resources that could go into making AGI.
There is nooooo way for companies to invest in actual innovation when they are throwing everything at this dead end.
Then you’re well aware of the massive power that AGI will bring to any nation that can harness it. And no, LLMs alone are not the path, and possibly not the path at all.
Making a better LLM isn’t the point of all this, it’s taking what they have and building on it until they create a true AGI.
Whoever gets there first, makes basically everything else obsolete in an instant.
In a world where the organisations that are blazing the trail are in private hands, this is very bad news for everyone who isn’t in the winning organisation.
That’s essentially the arms race: who gets to be king of the world.
The slim chance of it not being monumentally detrimental to humanity is basically tied to us abandoning capitalism wholesale and uniting the world, so I’m not holding my breath.
Edit: few downvotes on this, so check my other replies for clarity, if you still think I’m taking out my arse, comment and set me right. It’s Lemmy, the points don’t matter, I’d rather have a conversation. Plus read again if you somehow get the impression I’m advocating for any of this
Nobody’s making AGI anytime soon. LLMs do not have any of the baselines required for this. They’re expensive predictive text algorithms, more or less the same ones used in mobile keyboards, but upscaled to an absurd degree. Anyone truly worried about other companies or nations developing AGI has no idea how our current “AI” works. You’re never going to get there by building on them.
I’d like to believe too, but it doesn’t really track when you watch what these companies are actually doing.
Of course an LLM on its own isn’t going to become an AGI. Anyone with a braincell can see that. These orgs aren’t so high on their own farts that they ignore this.
Nearly all of the actual uses today aren’t just the LLM, but the tooling built on top of it, the LLM is the bit that you can plug into the past century of computing developments to enable much greater autonomy.
It’s true to say an LLM in isolation isn’t going to become AGI, but it’s also looking very likely that an AGI will feature an LLM as a key component.
That’s what’s happening in parallel to the model development, tooling and harnesses that make the overall system more capable. If it can be done by a computer (or by extension a sufficiently advanced robot), the LLM can do it too with a bit of integration work (which it is very able to do on its own today, with minimal steering). If you can test for something being correct in any way, that too can be ultimately hooked up to an LLM as another input to push it back onto the desired path when it veers off.
Frankly I’m starting to feel like for most people it’ll feel like it’s years off until the day it happens. I don’t see remotely enough people taking the risk seriously in time to do anything.
True AGI is not happening within the lifetime of anyone or anything alive today.
See my other comments
The AI “arms race” as you put it is absolutely capitalism at its core. Replace humans with shitty robots so they don’t have to continue paying wages to actual humans. Its just the the first person that makes it work will be able to set the rules for the ones that follow. Getting paid for those rules and making further entrenched in capitalism.
The frustrating part is that we could be on the precipice of an amazing time. We could be in a space where it makes sense to dump tons of resources into rapidly progressing automation because it would enable people to finally stop doing tedious labor.
But a combination of our inability to demand collective ownership of these systems and a similar disdain for social welfare means the prospect is instead terrifying. We need to continue to allow people to work cash registers for well below livable wages because otherwise they’ll starve.
There is an alternate reality where the end result of AI is that people are just free to live how they want, to socialize, to explore art and novel ideas within their passion, engage in social supports, etc. but instead we will continue to prop up the need for mind numbing and tedious labor out of a fear of homelessness because collectivism is scary and bad
I think we may very well be on the precipice of the world you imagine, or something like it. But the old world dies hard and takes effort to abolish. We didn’t get where we are because we were given what we have - we fought for it. I think we’re seeing the beginning stages of people demanding that the benefits of AI and automation flow to them, rather than to just the elite. Won’t be without pain, but I think we get there. Partly because we kind of have to. People get over their fear of socialism and collectivism really fast when they get desperate yet there’s people making huge piles of money off the automation that stole their job. I can’t say for sure what the future looks like, but I don’t think we stay locked here forever. To think so is to look at the situation during the first gilded age and say nothing can change. Well, it did and we got the progressive age.
The frustrating part is that we could be on the precipice of an amazing time.
Either the collapse of civilization or a literal uprising.
Seeing as we’re already heading down the nightmare route, it seems poor risk management.
I mean we passed that point decades ago, im not an expert but im pretty sure it’s literally been decades since we produced enough food for everyone on the planet to be obese and at least in the US I believe we have more empty houses than individual homeless people. AI overinvestment is another step in the wrong direction but it’s not the cause of any of our current problems.
I don’t think this is quite how the world works. The reason people need to work to survive is because we can’t survive if everyone stopped working. We have to make people work under threat of homelessness because if we didn’t there is too great a risk they wouldn’t work at all and that would eventually mean the collapse of society. How many people would quit their job tomorrow if they won the lottery? Sure some would find work doing something they preferred, but not all of them, and often the thing they prefer doing is not the job that actually needs doing the most. If it’s something they even are good at. Loads of people would love to be an actor, but how many actually have both the talent and the skills needed to do that?
In a society where most people actually don’t need to work because most work can be handled by machines without significant negative consequences things would change to be very different. People like to think rich people or politicians or kings control the world by themselves but the reality is there are always limits on what they can actually do. If you dick around too much even in an absolute monarchy you will be overthrown one way or another. Typically by your own military, underlings, or family, sometimes by revolution or insurrection. The same thing applies today to liberal democracy. In fact it applies even more so. Anyone who tried to kill off the working class as a whole would find themselves very quickly dead or dethroned one way or another.
It’s a tool, use it as you wish but you either have it or not.
Tool of your own destruction and a death knell of human logic, reasoning, and creativity.
No, thats not what I was talking about at all.
The thing, western governments fear is AI-powered terminators. They want the tech first, so they can win the war when someone attacks them. That is the arms race part.
The unemployment explosion is obviously also happening. But that’s actually a pretty good thing in the long run as a society with 90% unemployment and the need to work to live is absolutely unsustainable. AI will basically force the end of capitalism by increasing the system’s volatility until it adapts.
LLMs will never be able to be terminators. This is just an expensive exercise in futility.
I thought, LLMs would never become able to write code. And now, I use Claude Code as the always available senior on coke.
LLMs have a reliability problem. If that gets solved somehow, they can actually drive a worker bot - or a terminator.And the big money pits also don’t only do LLMs. Those just get all the press because they are usable by normal users right now. Of course, some of those money pits are just investor scams. It’s a fully corrupted society after all.
The “reliability problem” is a fundamental part of how they work. There is no solving it.
BTW, I wish you luck debugging all that vibe code when it stops working. Seriously.
Never say never
Except 90% of what people talk about when they refer to AI is LLMs which have no direct military applications other than vague productivity boost claims. You could say the same thing about sending kids to the mines, “our society is more productive sending kids to dig out coal instead of playing. If we don’t send our kids to the mines China will and then we’ll really be behind”.
… no I am talking about actual AI as a field as a whole, not just LLMs…
Yall forgetting about Boston dynamics or something?
We got fuckin guns strapped to the back of autonomous robot dogs that can run at over 60 km/h
We got autonomous drones with facial recognition that can fly through dense forests at 90 km/h+
The fuck you think Im talking about lol…
Boston dynamics isn’t building countless data centers and (poorly) replacing peoples’ jobs. OP’s comic is obviously about ChatGPT, Copilot, Gemini etc.
The biggest strength of LLMs is in processing a huge amount of text very quickly. I imagine that contributes a lot to military intelligence.
Again that’s a use case with entirely speculative applications at best. It can “process info”, but the answers it gives from them could be hallucinated. So if it’s analyzing nuclear threats, what if it “hallucinates” one? This isn’t Comcast ticket support, the military can’t afford LLMs making up bad info.
Implying that the only way to use LLMs is with 0 human oversight.
Once one country starts making nukes you cant stop everyone from following suit to protect themselves.
Except we did stop. We ended nuclear testing. We downsized our arsenal. We never deployed the high yield, neutron bombs, or other “tactical” variants in subsequent wars. And neither did the French or the Russians or the Chinese. Or even the Israelis.
Our nuclear program is derelict. It belongs in a museum. There’s an outstanding question as to how many of the bombs currently in circulation are duds.
Unlike with the F-35 or the Bradley Fighting Vehicle or the Predator Drone or even the Virginia class submarine, we’re just not putting any more money into nuclear proliferation and improved first strike capabilities like we were 60 years ago.
Mutually Asssured Destruction is an easily technolngically achievable goal, that’s why we don’t see much development past that point. You only need to exterminate an enemy’s population once.
The lesson learned by the world, at this point, is that nukes are the only way to guarantee your country isn’t invaded and that agreeing to unilateral nuclear disamament is downright idiotic. Expect more countries gettng nukes this century.
It would also apply to child labour and slavery. We may have outlawed it locally, but that doesn’t change the fact that companies who make use of it will be at an advantage, so we just ended up outsourcing it.
I find it ironic that every top comment author seems to feel the urge to point out why it’s actually different, but never question the point. I’m also sick of people telling me there’s no turning back, like, yeah, you do you, bro. My life is great without social networks, which are not going anywhere I guess.
lemmy isnt a social network?
Social media, IMO, is Facebook, Instagram, TikTok, etc. Lemmy, Reddit, and the like, are more akin to message boards, which I think of as being distinct. Reddit seems to be trying to transform itself, but I still use old.reddit, and once they get rid of that they get rid of me with t.
Well to be fair to the giant pieces of shit in positions of power all over the states they are trying to bring child labor back as well
I’d bet good money it’s the same people fighting to keep child marriage.
Except child and slave labour was cheap and profitable. AI is neither cheap nor profitable.
I am more mad about people saying “it’s improvng exponentially.” The rate of improvement is falling, if anything.
Bunch of people said it because sci-fi made them believe so, and then everyone else went along with it for some reason.
Either the exponent is 1/2 or people are just having shared delusions.
Or people don’t know what exponentially means
Ha the exponene might be half, but you’re right exponential is a huge oversimplification, even as the processing power increases exponentially, the resl world results increment much less quickly.
It’s comparing apples to oranges.
AI is software. We never stopped any software change before. Even heavily disliked and banned systems like crypto currency or vpns etc. still exist.
For the record I agree that AI needs more regulation and we could even force stop development of new models but LLMs will never be stopped in any meaningful way. You can take an open source model and run it today.
LLMs are here to stay until it’s replaced by other technology.
AI is software. We never stopped any software change before.
Good point, hard to stop something that has near zero cost of copying (see also ‘piracy’)
Which is why techbros are trying to put a moat around it with ‘datacentres’ . Problem is, as the tech advances, it keeps getting smaller. QWEN 3.6 27B can run fine on a 16GB video card and if you give it more time it’ll be as ‘smart’ as bigger models. Doesn’t have as much world knowledge as the bigs, but for many usecases that’s irrelevant.
Really, ‘datacentres’ are more about stealing compute from the masses so they can rent it back, with control.
QWEN 3.6 27B can run fine on a 16GB video card and if you give it more time it’ll be as ‘smart’ as bigger models.
As much as I’d like this to be true (don’t believe all the benchmarks), in reality, using e.g. gpt 5.5 is still a lot less pain in the ass, mostly has to do with more reprompting (gpt is just smarter, oneshots stuff more often) + a lot slower (on an RTX 3090 for reference).
I’ve tried using it for some time, but I think I’m faster writing (better, although that’s also true for gpt-5.5) code by hand, than using this (+ I need the valuable VRAM for other stuff, as I’m a graphics/shader programmer most of the time).
That said, it’s already fairly impressive how much progress these smaller models have made the last year, it’s usable, you can “vibe-code” at least simple stuff.
Eh, valid I do in fact agree and perhaps I exaggerated with 16GB, I do in fact have two,
As much as I’d like this to be true (don’t believe all the benchmarks), in reality, using e.g. gpt 5.5 is still a lot less pain in the ass, mostly has to do with more reprompting (gpt is just smarter, oneshots stuff more often) + a lot slower (on an RTX 3090 for reference).
You’re not wrong, but perhaps you are not giving them adequete time to be wrong. Oneshotting is not the be-all, right is, especially maintainably correct. I’ve found letting them fight over it useful.
You may have a good point here that changed my opinion on datacenter opposition. We definitely need more datacenter compute and always will but maybe making it more difficult can shift the market towards on device compute or smaller servers so opposing datacenters can be a net good outcome.
This article assumes the person in the first panel wouldn’t want the 3 panels to not still be the case.
As much as I hate ai/llm’s, here we’re conflating new technologies with bad practices. This is a fallacy.
However much we hate llm’s they definitely aren’t going away anytime soon. You can’t make laws or policies to make ideas/technologies go away.
You can’t make laws or policies to make ideas/technologies go away.
Yes you can. You write a bill saying AI data centers are banned. That’s it. That’s the bill.
About as effective as writing a bill saying war is banned.
https://futurism.com/data-centers-financial-bubble
The massive investments being made aren’t economically viable.
The current economy isn’t economically viable
Exactly. We burned all the bridges and laid bare some pretty obvious contradictions in the existing world order behind us! If this doesn’t work out, the world as we knew it is basically over.
Neither were the massive investments during the dot com bubble
Asbestos is already in all the buildings, we can’t remove it. All the cars already require leaded gasoline, we can’t unlead it.
Fun fact I didn’t know until recently: if you have a classic car that requires leaded gasoline, they actually sell lead substitute that you mix with modern unleaded gas
Isn’t the problem now the ethanol and not the octane?
I don’t know the exact mechanisms at play that make it necessary, but they sell the additive at automotive stores!
Agreed. How would someone even contain all the open source options that don’t require internet?
Nope. Good try, though.
“Leaded gasoline is already there, there’s no going back.”
A century later it isn’t gone yet, though for the most part we’ve replaced it with a slightly less lethal formulation
Reminds me of when the stupid “Web 3.0” made up by blockchain freaks was supposed to be the future. Not every technology will be as widespread as the internet. The internet facilitates communication across the entire world and offers many advantages over phone, mail, and other forms of communication.
The use cases and advantages are clear, even if there was an overly eager hype cycle in the 90s. AI might have some uses, but a clear advantage has not actually been established yet, nor have the legal challenges been ironed out. Remember that the current iteration of AI would not have been possible without breaking tons of IP law, slurping up as much data as possible.
What do you mean clear advantage has not actually been established?? Im getting weeks worth of stuff done in hours, like it’s self evident how powerful AI is. If you use it to make deepfake porn instead of making you better at your job, that’s a human choice. If you make a half decent effort, AI makes you an order of magnitude more productive. It’s pretty freaking amazing, like how the automobile was a massive improvement over a horse and buggy.
Agree AI is as overhyped as the internet in the late 90’s was. I also think AI or some descendent of it will likely be as ubiquitous as the Internet is now. There’s quite a few problems right now that AI just seems really well suited to solving, unlike the blockchain where it really only solved one sorta esoteric problem. I look at AI as being the bridge between the real world where things are fuzzy, rules are inconsistent, things don’t have clear cut answers, etc and the digital world where everything is precise and well defined. That’s not something that’s going away.
However, what I see happening with AI is much the same thing as what happened with the Internet. To use the Internet in the late 90’s was frustrating. The computers sucked, they were huge, they used a bunch of power, the connection was slow, connections dropped, they weren’t always on, they took quite some to establish, etc. It wasn’t till CPUs got good enough to be able to be battery powered and still render full websites (in other words, the key building block of a smartphone) that the Internet really became a ubiquitous thing for most people. Today’s AI uses way too much power, requires hardware that’s way too expensive, is less smart than people think it is, has problems learning, has problems with hallucinations, etc. What I see happening is the AI bubble crashes, like the dotcom crash, but then it comes back once the technology is really ready.
As far as law and IP go, the Internet often had lots of issues with that too. Lookup the origins of why we have Section 230. It’s still something we’re arguing over. We’ll figure out the legal issues. And IP law is broken, has been for a long time. It needs a revamp to bring it back to some sanity. I have no problem with AI breaking IP law. Much of that shouldn’t be under copyright anyway.
I feel like this comic is bait
id say the head over heels visceral hatred of ai is more common than any sort of praise. not sure who “all” means.
My wireborn girlfriend says otherwise! Try being less of a bigot!
lollll … hold my beer while i ponder my ways
That’s very big of you. Thank you.
![Stop telling me AI is the future [Still Vreni]](https://lemmy.world/pictrs/image/4bb4b8e1-242a-4f66-ae86-60cf2e04f996.jpeg){kind=link}