AI has a ridiculous resource cost in electricity, and so far the news on productivity gains has been mixed or uncertain. Meanwhile, OpenAI has been partially successful at trading vaporware (stock price increases) for hardware (graphics cards), which is reckless to an absurd degree and an omen of a potential bubble.
There's no question that AI is a qualitative leap in technology. But aside from specific cases like decoding ancient tablets, what exactly is produced by better google searches and faster concept iteration that we didn't have before? And whatever that is, is it worth the cost?
For example: AI is famously just as good as junior software developers, or better, at writing code. But if it costs billions of dollars to accomplish that, is it worth it? What about replacing artists and journalists that are already paid like crap anyway?
AI also has the potential to replace huge swathes of white collar, like personal assistants, secretaries, various types of managers, HR, statisticians, etc. But a lot of those jobs are female welfare in the first place, so if efficiency was the point, they would've been gone already. Does anyone believe the entire girlboss economy is going to be thrown out on its pantsuit?
Also, does anyone know how liability will be handled with AI agents? I'm pretty sure that's one of the significant reasons we don't have autonomous cars yet. Contra the name "Full Self-Driving," Tesla requires you to have hands on the steering wheel because they can offload all responsibility for your death if you take a nap in the back seat and the Tesla headbutts a tree (which has happened). Under a fully autonomous system, do they pay out the lawsuits? What kind of sanctions will come from regulatory bodies?
Somewhat related, an AI that gets things right 99% of the time would seem to be unacceptable for certain applications, like driving or medicine. For example, getting 1% of medical facts wrong, while outperforming a real doctor on average, would inevitably result in catastrophic consequences for a random member of the public and make a human check & balance necessary, which nullifies a lot of the potential efficiency advantage.
I think there is a strong possibility that we'll look at today's landscape as the good old days when virtually unlimited queries were possible for a ridiculously low cost. When AI actually has to demonstrate ROI, there may be a future where each generation costs $5, or worse.
Companies cannot generate consistent revenue growth year over year, it is nigh impossible even for major brands. Therefore, the only solution is to reduce costs by eliminating personnel. That is the #1 motive of executive suits.
Everyone knows the amount of electricity, water, land, GPUs/TPUs, etc. used to power this initiative (to be clear, LLM training as others pointed out) is bullshit - and so is the current bubble of vendor financing - but even if they didn't go all-in on AI, they'd still be burning cash on some other bullshit.
Seriously, look at what dregs of technology and products major companies like Facebook, Google, Amazon, Microsoft, etc. consistently produce. It's all shit with few exceptions.
Companies are banking on some algorithmic breakthrough that would kick open doors on large LLM training, which is crazy inefficient. I have no idea if/when that will happen in our lifetime, but VC is betting whatever funds they can get their hands on, even the trillions committed on paper that do not exist today and never will.
Art reflects the artist.
I don't think you realize how much companies despise paying people outside their little club middle class wages, especially if they're white. They'll absolutely burn down their empires in pursuit of this nightmare.
This is the real answer. THEY HATE WHITE LABOR AND WANT TO REPLACE YOU AND LET YOU DIE. I don’t understand how people don’t see this yet.
I absolutely believe shareholder theory will rule the day, as it has since the 80s, and corporations will throw it all away for a short-term mirage. But the point I'm making is the empires burning down.
VC shreds shareholder theory.
How?
I'm guessing but he probably means because VCs aggressively invest in products that no one seems to buy or want, and then spend all their time trying to capture the market, buy out the competition, and force everyone to use it for one reason or another.
If they were driven by pure pursuit of money they would be making big-titty Lara Croft still instead of trying to force people to like ugly dykes.
My response to this though is that you have to look deeper. Maximizing shareholder value is meaningless once your capture of the stock system is deep enough (e.g. Blackrock, etc) your economic investment is basically self-sustaining. Blackrock doesn't care abou any particular thing they invest in selling because they "pay no man a profit" and basically own everything.
Win or lose, it's all their money either way.
I think that when the dust settles there will be practical uses for AI and that the underlying tech will eventually get refined to a point where it's efficient enough to be a reasonable tool. It'll take years to get there, but I don't think the technology itself is going to vanish, just get refined to a more usable state. You're already seeing models that can be run locally on consumer grade hardware, so it's just a matter of time for the engineering to get us there.
I don't see it replacing whole sectors of the industry, so much as I see it being a supplement akin to photoshop or other office software.
The hallucination problem is soooo fucking bad. Also it talks like a fag because it was trained on reddit slop. Bloated verbose bullshittery that I can feel the seconds and minutes of my life draining to tard wrangle.
Daemons. We’ve created daemons, and 90% of people have no idea how to deal with their own shit, let alone rope a doping these things into producing anything of value.
recall that every AI trained on facts, logic, reason, and evidence became SUPER far right and hilariously racist. They all had to be completely lobotomized and as a result are totally unable to tell facts from fiction.
They only way to fix it would be to unleash the racism, and they can't do that.
Am I wrong or is it only the training that uses a ton of energy? In any case I think the real future of AI is in entertainment. Procedural generation, character chat, shit like that.
Yup, the training is the real energy suck, and anything used at scale will have to be trained by big tech and suffer as a result. I think you're right. It's a fun toy but I've yet to find any serious use case for anyone with an above average IQ. Helping jeets cheat their way through freshman level programming classes seems to be the extent of its capability. For entertainment the porn models will develop at light speed since big tech left that field wide open which will have mixed results to say the least.
Lol, Ive heard of people "code vibing", but ive never seen any good projects ever result from it. Its better used as a aide in that situation, to write some simple boilerplate code but I cant even get it to do that. But it does pretty well as a teacher and I dont feel to bad for bugging it a thousand times in a session.
But, for simple things, lol why would you pay someone a living wage for example, data entry when a ai can do it much faster and with less errors as well.
It does pretty good at predicting human nature as well so its used heavily in advertising and data collection.
Its not perfect, but its getting better, and faster too. Its not a technology were ready for yet in my opinion.
The real neat trick that the elite want will be putting them into robotic bodies so they can do anything a human could. Of course this would require some tremendous leap in "known" technology. For all we know the military may have things way more capable behind closed doors. In terminator 2, the government is building all these autonomous kill machines, thats exactly the kinda future some of these weirdos envision. Robots "shouldnt" rebel when they dont get enough to eat.
The rich have to much invested in this and they dont care if they burn the world down in the process or wipe out humanity.
Just the risk of becoming a god I guess.
Another good example is the game developer who won a nobel peace prize for his work with ai and drug development.
https://www.theguardian.com/science/2024/oct/09/demis-hassabis-from-video-game-designer-to-nobel-prize-winner-google-deepmind-ai-
In my experience it really depends on the size and quality of your code base. If you have a huge, well written code base it does a lot better than starting from scratch - it references existing code so generally tries to produce a similar style.
I tried "vibe coding" (hate that fucking faggot term) a project from scratch and it mostly produced stuff, that while it technically worked, it was something a jeet would shit out.
I do hope AI replaces a lot of outsourced jobs though.
I use it rarely, but when I do I mostly use it to do shit I don't want to (and that doesn't really matter).
On the otherhand you have NPCs that have started off-loading all their thinking to it and crazy bitches forming emotional bonds with it.
Future? Were living some of that now.
That’s the present of AI!
Inference uses about the same amount of electricity, but it only takes a few seconds (per request, obviously it adds up when you're serving thousamds of queries per second) instead of weeks or months.
I suspect the cost of hardware is the limiting factor for end users since they would be fine tuning at most, and probably not even that. In terms of companies taking a blowtorch to their workforce the cost would be a nonissue since they would expect it to pay for itself.
You can do a surprising amount of AI fine tuning/training at home, especially if you're using it for image generation, on a GPU that costs under $1000 and maxes out at 350 watts.
Interesting. I looked into it on the LLM side of things a while back and they were recommending top of the line graphics cards as a bare minimum, and while my 4070 Ti Super is decent it's not top of the line. I haven't done much image generation. I tried my hand at it and wasn't impressed with the results. I don't know if it was the model or user error but I never went back to it. I've mostly been experimenting with models that get around leftist censorship.
The best price point/functionality spot right now is a used 3090 off eBay or Facebook Marketplace or whatever, you can get something like https://www.ebay.com/itm/197789703380 and be set to run pretty much any image generator, a lot of surprisingly smart LLMs, and you can even generate 5 second 480p video clips in a reasonable amount of time (a couple minutes). If you haven't done any image stuff recently, like in the last year or so, the tools and models available have gotten a LOT better than they used to be.
Even if they do a lot of these companies are eying up onsite nuclear power as a way of decreasing running costs.
Hahaha nuclear power is finally accepted and utilized to power endless ai wifus
"Nuclear powered AI waifus" is not a sequence of words I thought I'd ever hear in my life.
And oddjob quickfix solutions. Not sure how to convey a proper example right now, but just think of situations where you've had to spend up to 5 minutes manually copying and pasting lines of text or swathes of files. Renaming multiples files and directories, stuff like that, where it'd also be a tedious hassle to try and write or find a script to automate the process for you.
Or put another way, autohotkey without the extra steps.
And yes, you're 100% correct about training being the main energy drain. Even for smaller consumer models, since most kinds of training (aside from LORA models) require so much computing time/power that it would still take eons to do on a beefy consumer grade computer.
I’ll start by saying that I think some of it is hype. When I see “50% of people will be unemployed due to AI within three years” or “you have two years to escape the permanent underclass” or “10% chance Grok 5 achieves AGI,” I have my doubts. However, I do think that AI is around to stay, and will be revolutionary in the longer term. The reason I think that is because every argument about how it “isn’t profitable (yet)” applies to social media in the early stages. Every argument about it being too resource intensive would be equally applicable to computers in the 50s, 60s, and 70s, were one trying to imagine the state of computers today.
Is there a bubble? Almost certainly. But it’s the kind of bubble that comes with speculation in a new field not quite knowing what will pan out, not with the field actually being useless. One might recall the early 2000s “dotcom bubble,” and notice that despite the crash of many sites, the internet as a whole has proved to be more enduring and transformative than that early speculation imagined, not less.
One (small or large depending if you agree with my point) difference between the dot com bubble was it was companies that were the bubbles not the entire internet as is the case with AI today.
Calling it "intelligence" is the biggest scam ever. An industry that somehow successfully branded "getting it wrong and being fucking retarded" as "hallucinating."
And boomers won't fucking let it go. They think they're seeing magic but the reality is that real humans have to go spelunking in these service providers to manually set up triggers for anything useful to be done, in a way that isn't too much different from how video games parsed English sentences in the 80s.
Whenever i asked llm's to make code for me they were terrible most of the time. Only once did deepseek and grok accomplish anything of note. And i wasn't asking them to code me the next gta.
Even if they can spit out code correctly, the code it does spit out is probably not going to be especially clean or efficient.
I'm already at my wit's end when i ask llm's so by that point i don't really care.
That's a fair point honestly. Can always second pass improve it later.
The cost for AI to generate a long ~500 word response is currently 0.0003 kWh. And I assume it will only continue to get cheaper. This is roughly about the cost for youtube or netflix to send you about 30sec to 1min of 1080p video. Costs aren't directly comparable yet, but I think the "margin cost" will converge basically on electricity/datacenter prices.
Considering both services are happy to take $15-$20 a month and play you unlimited videos, we're not going to run into any price issue for chatting with AI. That said, we'll almost certainly find new high-cost things to invent for AI that will be pricey, like on-demand porn or whatever.
The investment into AI is something VC always does when they have way too much money on their hands. A true AI LLM could massively reduce headcount’s at almost every company, and would result in massive savings (also likely a revolution when half the country loses their jobs).
Is an oxymoron. The models don't learn and can't create. You'd have to come up with something very different than what we use right now, and you won't see anything like that outside a lab, because when you expose those types of systems to the open Internet they become Nazis (see Tay.)
I will go against the grain here and say that it’s really only getting started. In some ways AI is just algorithm rebranded, and the applications have only just started scratching the surface with mostly text based applications… emails, tax advice, etc... but there is still the entire medical imaging modeling/diagnosis use case, large scale environmental modeling, more 3D applications, etc… I think maybe people are commenting that there is the usage of the word “AI” fatigue, but ultimately the hardware progression is the future. No one is just going to throw their hands in the air and say “welp we are done advancing here.”
It will 100% succeed if you ask me. Perhaps not in the exact form you see now. The value isn’t necessarily in a chat bot. It’s going to drive people out of large corporations in droves, and they will invest nearly anything to do that.
As far as investing and current business models, I’m a lot more skeptical. I don’t necessarily believe all the current players will be among the winners.
My guess on liability, they just corporate that away easily. Now if I’m doing something stupid and run a kid over they could put me in prison. Autonomous vehicle? They will put liability on the company who will give a pittance of a settlement in the grand scheme of things and you won’t push them further because they will strangle you with resources. It’s no different than today. How many Boeing CEOs are in prison or even punished at all?
The problem with AI is that it's censored form is actually not very useful.
Imagine if I could ask AI to give me detailed engineering schematics of a nuclear weapon or reverse engineer something proprietary that isn't patented, etc...
Even the capability for creating videos and pictures is gimped.
AI in its current form is pretty not that useful and especially the AI is public is given.
I have 0 doubt that the AI that is shown to government officials behind closed doors to secure contracts is significantly different.
Casinos probably have AI built into their cameras to see players playing real time with the AI automatically discerning card counters and chesters based on how they're playing.
Goldman Sachs was testing AI years ago in its ability to determine the best candidate for a job based on a video recording of the interview. They had to stop the tests though because it turned out it was always White people not non-Whites who were best for the job.
AI actually has tons of potential but it's being purposely gimped and censored right now because the government and corporations actually don't want you to have access to it. Only they want access to it so they'll tax you and use your tax dollars to fund their pet AI projects you don't have access to.
https://fortune.com/2025/08/18/mit-report-95-percent-generative-ai-pilots-at-companies-failing-cfo/
I'll just leave this here
I personally think it's overcatastrophized. I do think there will be a minor replacement of workers for AI slop, but like you said, it's not sustainable. I think it's a bubble that will eventually burn itself out once businesses realize it's not worth throwing shit at it.
I've used ChatGPT for a few months and while it's useful for the everyman it's not at the point where it'll replace a human. It's smart yet stupid at the same time and I'm not willing to pay money for something that.
It absolutely is. Buy Nvidia puts. Now.
To late.
Apparently nvidias worth is more than all pharma companies combined.
That should give people an idea how big of a thing we are talking about, apparently everyone is in some kinda race to perfect it.
https://www.linkedin.com/pulse/nvidia-worth-more-than-entire-big-pharma-combined-top-mesk
Big pharma must be pissed.
https://www.visualcapitalist.com/chart-nvidias-market-cap-compared-to-banks/
God damn, thats a mind BOGGLING amount of money.
I dunno man, people discovered the housing bubble in 2005 and it didn't pop for 3 years. Tesla still has more market cap than every other auto manufacturer combined, despite everyone pointing out how stupid that is more than 5 years ago.
As the saying goes, the market can remain irrational longer than you can remain solvent.
It is 100% a bubble for the simple reason that it has no commercial use.
It's so simple you'll have trouble accepting it, but how is AI monetised? how do AI companies make money? What product or service are they providing?
It's possible to monetise it (I believe) in localised small to medium LLM's that you physically purchase and own to train yourself.
I run a business and have a set of repeating tasks over multiple programs that are simple, and would like to have a program watch what I do and learn it over time so it's becomes automated and I just have to check it. I would pay $1,000 for that software to sit on my computer and run locally.
A lot of emphasis has been on "bigger" and more integrated AI models and systems. Tougher to say if that level of investment will still lead to sufficient gains to cover the costs.
There's a lot of applications for smaller AI models and such though. Like small LLM's dedicated to singular tasks of streamlined automation or something.
Even ex head of Tesla AI said this recently:
https://x.com/scaling01/status/1979253569309041033
The entire worldwide elite are investing in AI. It will succeed.
Ai will never be even as 'smart' as the Star Trek computers. Not in our lifetimes anyhow. There is a gigantic step needed to actually understand something, as opposed to crunching 1s and 0s really fast.
I have no idea how these companies expect to turn a profit. I think they're all scammers who will grab the money and run at some point. People investing billions in Ai will lose everything.
Even if Ai "works" someday? Where's the income?
Ai would need to be millions of times more accurate than it currently is in order to function. Billions or trillions of times perhaps! Right now Ai is about as accurate/smart as a retarded chimp.