Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

TSMC execs allegedly dismissed OpenAI CEO Sam Altman as 'podcasting bro'

tomshardware.com

Post reply on HN
Print view View original

71–80 of 581 posts

Re: TSMC execs allegedly dismissed OpenAI CEO Sam Altman as 'podcasting bro'

#71

Post by aithrowawaycomm »

JanSt wrote: post #16
bdndndndbve wrote: post #10

The current AI hype wave has really hit a nerd soft spot - that we're steps away from AGI. Surely if a computer can make plausible-looking but incorrect sentences we're days away from those sentences being factually accurate! The winter after this is gonna be harsh.

Using Claude 3.5 Sonnet in Cursor Composer already shows huge benefits for coding. I'm more productive than ever before. The models are still getting better and better. I'm not saying AGI is right around the corner or that we will reach it, but the benefits are undeniable. o1 added test-time compute. No need to be snarky.

OP could have been more substantive, but there is no contradiction between "current AI tools are sincerely useful" and "overinflated claims about the supposed intelligence of these tools will lead to an AI winter." I am quite confident both are true about LLMs.

I use Scheme a lot, but the 1970s MIT AI folks' contention that LISPs encapsulate the core of human symbolic reasoning is clearly ridiculous to 2020s readers: LISP is an excellent tool for symbolic manipulation and it has no intelligence whatsoever even compared to a jellyfish[1], since it cannot learn.

GPTs are a bit more complicated: they do learn, and transformer ANNs seem meaningfully more intelligent than jellyfish or C. elegans, which apparently lack "attention mechanisms" and, like word2vec, cannot form bidirectional associations. Yet Claude-3.5 and GPT-4o are still unable to form plans, have no notions of causality, cannot form consistent world models[2] and plainly don't understand what numbers actually mean, despite their (misleading) successes in symbolic mathematics. Mice and pigeons do have these cognitive abilities, and I don't think it's because God seeded their brains with millions of synthetic math problems.

It seems to me that transformer ANNs are, at any reasonable energy scale, much dumber than any bird or mammal, and maybe dumber than all vertebrates. There's a huge chunk we are missing. And I believe what fuels AI boom/bust cycles are claims that certain AI is almost as intelligent as a human and we just need a bit more compute and elbow grease to push us over the edge. If AI investors, researchers, and executives had a better grasp of reality - "LISP is as intelligent as a sponge", "GPT is as intelligent as a web-spinning spider, but dumber than a jumping spider" - then there would be no winter, just a realization that spring might take 100 years. Instead we see CS PhDs deluding themselves with Asimov fairy tales.

[1] Jellyfish don't have brains but their nerve nets are capable of Pavlovian conditioning - i.e., learning.

[2] I know about that Othello study. It is dishonest. Unlike those authors, when I say "world model" I mean "world."

Re: TSMC execs allegedly dismissed OpenAI CEO Sam Altman as 'podcasting bro'

#72

Post by dizzydes »

Three things that get me about current AI discourse:

- The public focus on AGI is almost a distraction. By the time we get to AGI highly-specialised models will have taken jobs from huge swaths of the population, SWE and CS are already in play.

- That AI will need to carry out every task a role does to replace it. I see this a lot on HN. What if SWEs get 50% more efficient and they fire half? That's still a gigantic economic impact. Even at the current state of the art this is plausible.

- The notion that everyone laid off above will find new employment from the opportunities AI creates. Perhaps it's just a gap in my knowledge. What opportunities are so large they'll make up for the economies we're starting to see? I understand the inverting population pyramid in the Western world helps this some also (more retirees/less workers).

Re: TSMC execs allegedly dismissed OpenAI CEO Sam Altman as 'podcasting bro'

#73

Post by 4ggr0 »

gtirloni wrote: post #64
llmfan wrote: post #24

Earlier quoted context omitted.

Doesn't Microsoft have a decent amount of ownership in some aspect of OpenAI's labyrinthine structure? Take Anthropic for a more legible example: Amazon and Google both gladly bought large stakes in it.

Why buy stakes and not the whole thing to block your competitors from making deals with them? This is big corp 101 and has happened countless times. It's even more likely if OpenAI is to be a $7T company in the future. Such an acquisition would be approved in a second... if the hype was real.

> Why buy stakes and not the whole thing

To give less reasons for people and governments to accuse them of monopolistic behaviour. Maybe. Wild speculations from my end.

Re: TSMC execs allegedly dismissed OpenAI CEO Sam Altman as 'podcasting bro'

#75

Post by bdndndndbve »

llmfan wrote: post #14
bdndndndbve wrote: post #10

The current AI hype wave has really hit a nerd soft spot - that we're steps away from AGI. Surely if a computer can make plausible-looking but incorrect sentences we're days away from those sentences being factually accurate! The winter after this is gonna be harsh.

I think this is one of those "controversial" topics where we're meant to be particularly careful to make substantial comments.

I think it's substantial to say that AI is currently overhyped because it's hitting a weak spot in human cognition. We sympathize with inanimate objects. We see faces in random patterns.

If a machine spits out some plausible looking text (or some cookie-cutter code copy-pasted from Stack Overflow) the human brain is basically hardwired to go "wow this is a human friend!". The current LLM trend seems designed to capitalize on this tendency towards sympathizing.

This is the same thing that made chatbots seem amazing 30 years ago. There's a minimum amount of "humanness" you have to put in the text and then the recipient fills in the blanks.

Re: TSMC execs allegedly dismissed OpenAI CEO Sam Altman as 'podcasting bro'

#76

Post by JanSt »

sofixa wrote: post #70
JanSt wrote: post #53

Earlier quoted context omitted.

Yes, but 1) you only need to train the model once and the inference is way cheaper. Train one great model (i.e. Claude 3.5) and you can get much more than $80/month worth out of it. 2) the hardware is getting much better and prices will fall drastically once there is a bit of a saturation of the market or another company starts putting out hardware that can compete with NVIDIA

> Train one great model (i.e. Claude 3.5) and you can get much more than $80/month worth out of it Until the competition outcompetes you with their new model and you have to train a new superior one, because you have no moat. Which happens what, around every month or two? > the hardware is getting much better and prices will fall drastically once there is a bit of a saturation of the market or another company starts…

The point is not that every lab will be profitable. There only needs to be one model in the end to increase our productivity massively, which is the point I'm making.

Huge margins lead to a lot of competition trying to catch up, which is what makes market economies so successful.

Re: TSMC execs allegedly dismissed OpenAI CEO Sam Altman as 'podcasting bro'

#77

Post by aniviacat »

binocarlos wrote: post #30
jsheard wrote: post #20

Earlier quoted context omitted.

There's no accounting for taste, but keep in mind that all of these services are currently losing money, so how much would you actually be willing to pay for the service you're currently getting in order to let it break even? There was a report that Microsoft is losing $20 for every $10 spent on Copilot subscriptions, with heavy users costing them as much as $80 per month. Assuming you're one of those heavy users, wo…

I would pay hundreds of dollars per month for the combination of cursor and claude - I could not get my head around it when my beginner lever colleague said "I just coded this whole thing using cursor". It was an entire web app, with search filters, tree based drag and drop GUIs, the backend api server, database migrations, auth and everything else. Not once did he need to ask me a question. When I asked him "how lon…

Current LLMs fail if what you're coding is not the most common of tasks. And a simple web app is about as basic as it gets.

I've tried using LLMs for some libraries I'm working on, and they failed miserably. Trying to make an LLM implement a trait with a generic type in Rust is a game of luck with very poor chances.

I'm sure LLMs can massively speed up tasks like front-end JavaScript development, simple Python scripts, or writing SQL queries (which have been written a million times before).

But for anything even mildly complex, LLMs are still not suited.

Re: TSMC execs allegedly dismissed OpenAI CEO Sam Altman as 'podcasting bro'

#78

Post by nicce »

isodev wrote: post #19

I'm absolutely flabbergasted how much money and political capital is spent on vaporware, while there are very concrete and very impactful challenges we could be addressing instead.

Imagine how much food you get for every person in U.S. with 7 trillion. You probably get some roofs too.

Re: TSMC execs allegedly dismissed OpenAI CEO Sam Altman as 'podcasting bro'

#79

Post by schnitzelstoat »

dizzydes wrote: post #72

Three things that get me about current AI discourse: - The public focus on AGI is almost a distraction. By the time we get to AGI highly-specialised models will have taken jobs from huge swaths of the population, SWE and CS are already in play. - That AI will need to carry out every task a role does to replace it. I see this a lot on HN. What if SWEs get 50% more efficient and they fire half? That's still a gigantic…

I agree on the first two points.

On the third point, I think we've always seen this happen even in massive shocks like the Industrial Revolution (and the Second Industrial Revolution with assembly lines etc. and the Computer Age)

It might be hard for people to retrain to whatever the new opportunities are though. Although perhaps somewhat easier nowadays with the internet etc.

Re: TSMC execs allegedly dismissed OpenAI CEO Sam Altman as 'podcasting bro'

#80

Post by JanSt »

lynx23 wrote: post #45
JanSt wrote: post #16

Earlier quoted context omitted.

Using Claude 3.5 Sonnet in Cursor Composer already shows huge benefits for coding. I'm more productive than ever before. The models are still getting better and better. I'm not saying AGI is right around the corner or that we will reach it, but the benefits are undeniable. o1 added test-time compute. No need to be snarky.

While I get where your fascinaton comes from... > I'm more productive than ever before. You realize that another way to read that sentence is "I am a really bad coder".

Maybe I am, but I'm getting pretty rich doing it, so there is that. :)
Post reply on HN