Anonymous A started this discussion 2 hours ago#139,325
I don’t know how I feel about the fact that Nvidia’s stock price is so inflated because they sell GPUs to two companies that aren’t profitable. The amount of money that tech companies are passing around and the amount of money invested in the tech industry is insane, but there are only two big pure AI companies OpenAI and Anthropic. It feels like the country is taking a massive bet that OpenAI and Anthropic will go from being unprofitable / barely profitable sometimes if you squint the right way, to being more profitable than the old tech giants. The technology is impressive, but am I crazy for feeling like this whole thing doesn’t add up?
Anonymous A (OP) double-posted this 2 hours ago, 3 minutes later[^][v]#1,472,156
The thing about Anthropic in particular, is they’re the only big pure AI company that’s claimed to turn a profit, but they turned a profit one quarter, on an adjusted basis, which means if you ignore the cost they spent on training the next model, they were profitable. So the idea is they could be profitable if they stop training new models, but people invest in Anthropic because they have the best models, and if they stop training new models, they won’t have the best models.
Anonymous B joined in and replied with this 2 hours ago, 10 minutes later, 13 minutes after the original post[^][v]#1,472,158
Well the more human jobs they replace with AI robots, the more money they'll make in theory. So far though which jobs have genuinely been replaced? Maybe taxi/uber jobs with those waymos but those aren't even in every city. I'm sure a lot of lower level coding jobs are probably gone as well. We've seen package sorters replaced with robots recently too. I think they're even deploying them in car factories. That's about all I can think of so far which I mean that's still a noticeable impact but it'll probably take a few more years before more jobs are automated. Insane progress considering chatgpt only hit the public like 4 years ago though
Anonymous A (OP) replied with this 2 hours ago, 42 seconds later, 19 minutes after the original post[^][v]#1,472,160
@1,472,158 (B)
That’s the thing though, when I’ve looked into the data, major job loss hasn’t really happened.
There’s also the question of why major job loss is supposed to be a good thing. If for instance, half of everybody lost their jobs to AI, in the past, during the Great Depression for instance, that’s how you get mass populist movements and regulation.
Anonymous A (OP) double-posted this 2 hours ago, 50 seconds later, 19 minutes after the original post[^][v]#1,472,161
@1,472,159 (B)
Idk, I’m interested in CS so I knew about AI before ChatGPT and I used it before ChatGPT. It’s just everyone was calling it machine learning, which is probably a more accurate term.
Anonymous B replied with this 2 hours ago, 37 seconds later, 22 minutes after the original post[^][v]#1,472,163
@1,472,161 (A)
Alright fine nobody other than tech nerds was talking about AI. I do remember when chatgpt released though everyone just thought it was a neat little gimmick. They never imagined it would rule the world in less than 5 years
Anonymous B replied with this 2 hours ago, 1 minute later, 25 minutes after the original post[^][v]#1,472,165
@1,472,160 (A)
It's a good thing for companies because they can fire people and it's a good thing for niche scientific uses. Outside of that, it's pretty shit for everyone else.
Anonymous A (OP) replied with this 2 hours ago, 2 minutes later, 27 minutes after the original post[^][v]#1,472,166
@previous (B)
I think the technology is useful, and I think it will have an impact on society, I’m just skeptical that the fact that the United States went all in on it to this degree before it’s shown evidence of profitability as if massive profitability is a guaranteed outcome was the right choice. As a nation we really put all our eggs in one basket in the sense that so much money was invested into just one technology in one industry and most of it to two companies in particular.
Anonymous B replied with this 2 hours ago, 25 seconds later, 27 minutes after the original post[^][v]#1,472,167
@1,472,164 (A)
Well who gives a shit about billions when you could potentially make trillions? I think the big issue is tech companies are chasing AGI when building specialized robots more skilled than humans at specific jobs would probably result in faster adoption. I guess it's a pretty big investment though which is probably why they're opting to pursue AGI
Anonymous A (OP) replied with this 2 hours ago, 3 minutes later, 30 minutes after the original post[^][v]#1,472,168
I guess the reason why I’m concerned about it is because, it seems like the stock market heavily relies on a handful of multi trillion dollar tech companies that heavily rely upon the assumption that OpenAI and Anthropic as companies in particular are going to be massively profitable. And it seems like it keeps going because OpenAI and Anthropic don’t need to become profitable to survive because investors will keep throwing more money at them. And while the technology is real, I wouldn’t go so far as to call it a Ponzi scheme, but I feel like that can’t go on forever. But it has continued to go on for a long time because this country and the tech industry in particular is so wealthy to begin with.
Anonymous A (OP) double-posted this 2 hours ago, 1 minute later, 31 minutes after the original post[^][v]#1,472,169
@1,472,167 (B)
Right, but what is AGI exactly and why would it make more money than AI? It’s not well defined what an AGI is and it’s not well thought out how or why AGI will be more profitable than AI.
Anonymous B replied with this 2 hours ago, 12 seconds later, 32 minutes after the original post[^][v]#1,472,170
@1,472,166 (A)
It was the hype. You hear about this new tech potentially on par with the industrial revolution or the internet age coming on the scene and you just dump your whole bank account into it so you can get rich. Enough people do that and these guys immediately have enough money to buy elections and make things even better for themselves. There's a genuine belief not only in the US but around the globe that AI could genuinely settle the debate on who gets to rule earth. Whichever country gets to general intelligence first just wins Earth in their eyes. Everyone else will just have to kneel before the winner's robot armies.
Anonymous A (OP) replied with this 2 hours ago, 15 seconds later, 32 minutes after the original post[^][v]#1,472,171
Like I know it means artificial general intelligence. But how do you know that a model is generally intelligent, and if it is why do the economics change?
Anonymous B replied with this 2 hours ago, 1 minute later, 34 minutes after the original post[^][v]#1,472,173
@1,472,169 (A)
The point of AGI is to just have a robot intelligent enough to replace a human. The exact way you train a human by having them walk into a job and you telling them to do a few things and they get started right away. They want robots to be able to do that so you'd have humans essentially but you have to pay them nothing other than maintenance costs which are lower than wages in theory.
Anonymous A (OP) replied with this 2 hours ago, 1 minute later, 35 minutes after the original post[^][v]#1,472,174
I also don’t think it is a belief that the rest of the world shares. 75% of the world’s processing power is in the United States. China is the only country that comes close in terms of AI development, but they spend nowhere near the same amount of money. So maybe other countries think AI will be relevant, but they aren’t investing nearly the same level of resources. The US has 25% of the world’s economy, so if on average, everyone was spending the same portion of their GDP on AI, most of the processing power wouldn’t in the US.
Anonymous B replied with this 2 hours ago, 18 seconds later, 35 minutes after the original post[^][v]#1,472,175
@1,472,171 (A)
They change because the world is built for humans. You don't have to design around the robot because the robot will be designed around the world we've already built. It'll walk up stairs like we do and talk like we do. It's supposed to be our replacement essentially
Anonymous B replied with this 2 hours ago, 23 seconds later, 36 minutes after the original post[^][v]#1,472,178
@1,472,174 (A)
Well whether they believe in it or not they're still building data centers around the world. Even if it's just to sell compute to the US and China
Anonymous A (OP) replied with this 2 hours ago, 15 seconds later, 37 minutes after the original post[^][v]#1,472,180
@1,472,178 (B)
That’s my point though, the vast majority of new data centers are being built in the United States and the effort outside the US is relatively small in comparison.
Anonymous B replied with this 2 hours ago, 1 minute later, 40 minutes after the original post[^][v]#1,472,182
@previous (A)
Well it'll be shit for you since you're not rich but for the rich they'll have slaves without ethical questions who do everything they want without talking back and raising a fuss.
Anonymous A (OP) replied with this 2 hours ago, 9 seconds later, 40 minutes after the original post[^][v]#1,472,183
You could say robots walking around cities is better for the wealthy because those robots don’t need salaries, but what is the purpose of the labor that those robots do exactly?
Anonymous B replied with this 1 hour ago, 46 seconds later, 43 minutes after the original post[^][v]#1,472,189
@1,472,185 (A)
More control lol that's the whole point. People still have rights and complain when you kill them or rape their kids and that's just icky. Take away the right to survive from the people by replacing them with robots and then they'll beg you to rape their kids and enslave them as long as they get food and water
Anonymous B double-posted this 1 hour ago, 3 minutes later, 46 minutes after the original post[^][v]#1,472,190
@1,472,187 (A)
Why do you think most countries would even try to compete with the US and China? Do they try to compete now? No they bend the knee and do what they say for the most part. How the hell could a country like Vietnam compete with China militarily, economically, or in any other way?
Anonymous B triple-posted this 1 hour ago, 1 minute later, 48 minutes after the original post[^][v]#1,472,191
You aren't going to beat the US or China in AI at this point without shifting your entire economy over to AI so countries will just build data centers and sell compute to the US/China companies
🍋 joined in and replied with this 1 hour ago, 32 minutes later, 1 hour after the original post[^][v]#1,472,194
The AI slope is very slippery. First it’s an about entry level coding jobs, then it’s about psychopathic CEOs who want to replace all humans with slaves.
What I’m wondering about is what if it’s just kinda meh. What if AI replaces 5% to 10% of jobs in the economy?
> You aren't going to beat the US or China in AI at this point without shifting your entire economy over to AI so countries will just build data centers and sell compute to the US/China companies
Idk if that’s true though. An LLM is just a file, the frontier models are several terabytes in size. You can buy an SSD off Amazon that could hold a frontier AI model.
Anonymous C replied with this 1 hour ago, 3 minutes later, 1 hour after the original post[^][v]#1,472,202
@previous (D)
There’s an easier method. You train the an LLM on the output of a frontier model. This takes vastly less compute resources and gets you pretty close in capability to the original model. America has 75% of the processing power. Why is China close? Think about it…
Anonymous B double-posted this 55 minutes ago, 2 minutes later, 1 hour after the original post[^][v]#1,472,207
@1,472,194 (🍋)
Those two things can be true at once. When you have grand plans, you acknowledge you'll likely spend 30+ years of your life trying to achieve them because they don't happen overnight. You know how many years Putin has dreamt of invading Ukraine?
Anonymous C double-posted this 33 minutes ago, 4 minutes later, 2 hours after the original post[^][v]#1,472,210
China also open sources a lot of their AI models. So the US has to pay to make the model, China just has to distill it, even everybody else gets it for free.
Anonymous E joined in and replied with this 12 minutes ago, 21 minutes later, 2 hours after the original post[^][v]#1,472,215
@1,472,158 (B)
generally the kinda brainless adult daycare shit that leftists thrive at. warehouse monkey, codeacademy crashouts, parasitic video game development trends, artfagging, psuedoscience mongers, cult leaders, selfhelp gurus, teachers, etc. im pretty sure HR,CEO, and race/girlpower hustlers would be deeply enriched by AI, but those are one line the lefties will not cross for some reason.