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

GPT-5 is behind schedule

wsj.com

Post reply on HN
Print view View original

551–560 of 1001 posts

Re: GPT-5 is behind schedule

#551

Post by ciconia »

t_serpico wrote: post #208

One fundamental challenge to me is that if each training run because more and more expensive, the time it takes it to learn what works/doesn't work widens. Half a billion dollars for training a model is already nuts, but if it takes 100 iterations to perfect it, you've cumulatively spent 50 billion dollars... Smaller models may actually be where rapid innovation continues simply because of tighter feedback loops. O3…

When you think about it it's astounding how much energy this technology consumes versus a human brain which runs at ~20W [1].

[1] https://hypertextbook.com/facts/2001/JacquelineLing.shtml

Re: GPT-5 is behind schedule

#552

Post by jstummbillig »

sollewitt wrote: post #493
atonse wrote: post #457

Earlier quoted context omitted.

Same here. The ability to “talk to an expert” about any topic I’m curious about and ask very specific questions has been invaluable to me. It reminds me of being a kid and asking my grandpa a million questions, like how light bulbs worked, or what was inside his radio, or how do we have day and night. And before anyone talks about accuracy or hallucinations, these conversations usually are treated as starting off poi…

LLMs suffer from the "Igon Value Problem" https://rationalwiki.org/wiki/Igon_Value_Problem Similar to reading a pop sci book, you're getting an entertainment from a thing with no actual understanding of the source material rather than an education.

You just state this as if it was obviously true, but I don't see how. Why is using LLM like reading a pop sci book and not like reading a history book? Or even less like either, because you have to continually ask questions to get anything?

Re: GPT-5 is behind schedule

#553

Post by rchaves »

Nah it's just a marketing problem, "GPT" and "ChatGPT" names is the biggest asset OpenAI has, people have expectations so high for GPT-5 that they cannot burn this name unless it's something truly majestic, bordering AGI at the very least. Until they are confident enough that people will be blown off by it, it's better to continue building up the hype

Re: GPT-5 is behind schedule

#554

Post by astrange »

ijustlovemath wrote: post #147
RedNifre wrote: post #111

Earlier quoted context omitted.

But who rates the synthetic data? If it is humans, I can understand that this is another way to get human knowledge into it, but if it's rated by AI, isn't it just a convoluted way of copying the rating AI's knowledge?

This is the bit I've never understood about training AI on its own output; won't you just regress to the mean?

It's not trained on its own output. You can generate infinite correctly worked out math traces and train on those.

Re: GPT-5 is behind schedule

#555

Post by Fargren »

jstummbillig wrote: post #552
sollewitt wrote: post #493

Earlier quoted context omitted.

LLMs suffer from the "Igon Value Problem" https://rationalwiki.org/wiki/Igon_Value_Problem Similar to reading a pop sci book, you're getting an entertainment from a thing with no actual understanding of the source material rather than an education.

You just state this as if it was obviously true, but I don't see how. Why is using LLM like reading a pop sci book and not like reading a history book? Or even less like either, because you have to continually ask questions to get anything?

A history book is written by someone who knows the topic, and then reviewed by more people who also know the topic, and then it's out there where people can read it and criticize it if it's wrong about the topic.

A question asked to an AI is not reviewed by anyone, and it's ephemeral. The AI can answer "yes" today, and "no" tomorrow, so it's not possible to build a consensus on whether it answers specific questions correctly.

Re: GPT-5 is behind schedule

#556

Post by bradchris »

croes wrote: post #500
synergy20 wrote: post #481

Earlier quoted context omitted.

so do teachers and books, in the future we need have multiple variants to cross check

Cross check against what? AI generated texts will flood the internet and burry the real knowledge just like SEO did before. But this time the fake knowledge will be less obvious and harder to check.

If that turns out to be true, the it looks like AI just gave universities a new reason for being.

What a shift from twenty years ago when optimism over “information superhighways” on the “world wide web” would end knowledge gatekeeping and educate the masses, to now— worries of AI slop and finely tuned ML algorithms frying older and younger generations’ brains, while information of human value gets buried, siloed, and paywalled, with no way to verify anything at all.

Re: GPT-5 is behind schedule

#557

Post by FergusArgyll »

superultra wrote: post #489
croes wrote: post #479

Earlier quoted context omitted.

How do you know the answers are correct? More than once I got eloquent answer that are completely wrong.

I give AI a “water cooler chat” level of veracity, which means it’s about as true as chatting with a coworker at a water cooler when that used to happen. Which is to say if I just need to file the information away as a “huh” it’s fine, but if I need to act on it or cite it, I need to do deeper research.

Yes, so often I see/hear people asking "But how can you trust it?!"

I'm asking it a question about social dynamics in the USSR, what's the worst thing that'll happen?! I'll get the wrong impression?

What are people using this for? are you building a nuclear reactor where every mistake is catastrophic?

Almost none of my interactions with LLMs "Matter", they are things I'm curious about, if 10 out of 100 things I learnt from it are false, then I learned 90 new things. And these are things which mostly I'd have no way to learn about otherwise (without spending significant money on books/classes etc.)

Re: GPT-5 is behind schedule

#558

Post by josu »

bloppe wrote: post #485
lumost wrote: post #461

Earlier quoted context omitted.

At this point it’s quite likely that they could pivot and just be the chatgpt company. I’ve found chatgpt-4o with web search and plugins to be more useful than o1 for most tasks. It’s possible we’re nearing the end of the LLM race, but I doubt that’s the end of the AI story this decade, or OpenAI.

Ya I think they probably will, but "the chatgpt company" is not worth 157B. It might not even be worth 1B.

It has replaced ~50% of my Google searches.

Re: GPT-5 is behind schedule

#559

Post by DanHulton »

patcon wrote: post #512
croes wrote: post #479

Earlier quoted context omitted.

How do you know the answers are correct? More than once I got eloquent answer that are completely wrong.

There's something here that I feel is pretty deep, though offensive for some minds: What is the actual consequence of being wrong? Of not getting right the base reality of a situation? Usually, stasis is the enemy that is much great than false information. If people with 90% truth can take a step forward in the world, even if they mistakenly think they have 100% truth, what does it matter? They're learning more and a…

> “My father once told me that respect for truth comes close to being the basis for all morality. 'Something cannot emerge from nothing,' he said. This is profound thinking if you understand how unstable 'the truth' can be.”

Frank Herbert, Dune

Re: GPT-5 is behind schedule

#560

Post by scott_w »

wilg wrote: post #548
scott_w wrote: post #542

Earlier quoted context omitted.

A Scientific American issue where the authors have no idea that they don’t know a topic so just completely make up the content, including the sources. At least magazine authors are reading the sources before misunderstanding the content (or asking the authors what the research means). I don’t even trust the summaries after watching LLMs think we have meetings about my boss’s cat just because I mentioned it once as sh…

Its good to not trust it but that's not the same as it having no idea . There is a lot of value in being close for many tasks!

I think it’s a very dangerous place to be in an area you’re not familiar with. I can read Python code and figure out if it’s what I want or not. I couldn’t read an article about physics and tell you what’s accurate and what’s not.

Legal Eagle has a great video on how ChatGPT was used to present a legal argument, including made up case references! Stuff like this is why I’m wary to rely on it in areas outside of my expertise.

Post reply on HN