GPT-5 is behind schedule • Hacker News Board

Re: GPT-5 is behind schedule

#881

Post by superluserdo » Mon, Dec 23, 2024, 7:29 PM UTC

wing-_-nuts wrote: post #767

jpablo wrote: post #753

Earlier quoted context omitted.

They aren't blocking anything. They are just asking nicely not to be crawled. Given that AI companies haven't cared a single bit about ripping of other's peoples data I don't see why they would care now.

A number of sites have started outright blocking any traffic that looks remotely suspicious. This has made browsing with a vpn a bit of a pain.

I wish I could go back to the days of doing almost anything at all without having to tell a server what a motorbike or traffic light is.

Re: GPT-5 is behind schedule

#882

Post by cshores » Mon, Dec 23, 2024, 7:30 PM UTC

pphysch wrote: post #840

cshores wrote: post #815

Earlier quoted context omitted.

It ultimately doesn't matter because a fairly current snapshot of all of the world's information is already housed in their data lakes. The next stage for AI training is to generate synthetic data either by other AI or by simulations to further train on as human generated content can only go so far.

How is synthetic data supposed to work? Broadly speaking, ML is about extracting signal from noisy data and learning the subtle patterns. If there is untapped signal in existing datasets, then learning processes should be improved. It does not follow that there should be a separate economic step where someone produces "synthetic data" from the real data, and then we treat the fake data as real data. From a scientific…

As others have mentioned, Tesla is already implementing similar advancements. More broadly, a new AI framework called Genesis has emerged, capable of training robots in just minutes using purely synthetic data. It generates a virtual environment for the robot to "perceive" and train within, even though this environment doesn't physically exist. This is just one example. Another could involve an AI specifically trained to diagnose illnesses based on genetic information in DNA. The insights gained from this virtual scientist could then cross-pollinate with other AIs, enhancing their training and capabilities as well.

Re: GPT-5 is behind schedule

#883

Post by asdff » Mon, Dec 23, 2024, 7:33 PM UTC

anshulbhide wrote: post #510

I don't really care. I had to come up with a proposal to build a new R&D centre recently. To provide context on what our company does, I wrote a web scraper to scrape our own website (faster than going to IT) using Replit Agent and then fed that into O1 as context to come up with the proposal. In less than an hour. There is no going back.

If you managed to get the content by scraping the home website, did you really need to build a scraper to collect this and feed it into input for a model and have it spit out a proposal that you probably proofread anyhow? Or, could you have simply taken your own knowledge of the company, regurgitated some generic corporate speak yourself on top of that, and came up with more or less the same end product in maybe even less time?

Re: GPT-5 is behind schedule

#884

Post by efilife » Mon, Dec 23, 2024, 7:35 PM UTC

bluelightning2k wrote: post #325

simonw wrote: post #146

Earlier quoted context omitted.

No, I'm complaining that just because GPT-4 is called GPT-4 doesn't mean it's the fourth LLM from OpenAI. Off the top of my head: GPT-2, Codex, GPT-3 in three different flavors (babbage, curie, davinci), GPT-3.5. Suggesting that GPT-4 was "fourth" simply isn't credible. Just the other day they announced a jump from o1 to o3, skipping o2 purely because it's already the name of a major telecommunications brand in Europ…

There's at least 4 major releases just in GPT4. GPT4, GPT4T, Gpt4o-Mini, GPT4o,

There are releases. Releases is plural

Re: GPT-5 is behind schedule

#885

Post by asdff » Mon, Dec 23, 2024, 7:36 PM UTC

LightBug1 wrote: post #783

I just want to say something random. ChatGPT cited and English newspaper called 'The Sun' in the answer to a query I made this moring. I found that amusing and figured then that we might already be beyond peak AI! If AI is trained on sources that include The Sun, it's going to end up being garbage and/or constantly needing to be fact checked.

About once a week the google ai summary completely flubs the numbers of something I query on google. If you did that on stackoverflow you’d be flamed and put in your place with the right answer. No one is flaming the ai, it just became the source of truth one day with billions invested.

Re: GPT-5 is behind schedule

#886

Post by alach11 » Mon, Dec 23, 2024, 7:37 PM UTC

marcus_holmes wrote: post #333

danpalmer wrote: post #320

Earlier quoted context omitted.

Interesting idea. The concept of The Singularity would seem to go against this, but I do feel that seems unlikely and that a gradual transition is more likely. However, is that AGI, or is it just ubiquitous AI? I’d agree that, like self driving cars, we’re going to experience a decade or so transition into AI being everywhere. But is it AGI when we get there? I think it’ll be many different systems each providing an…

The Singularity is caused by AI being able to design better AI. There's probably some AI startup trying to work on this at the moment, but I don't think any of the big boys are working on how to get an LLM to design a better LLM. I still like the analogy of this being a really smart lawn mower, and we're expecting it to suddenly be able to do the laundry because it gets so smart at mowing the lawn. I think LLMs are g…

> I don't think any of the big boys are working on how to get an LLM to design a better LLM

Not sure if you count this as "working on it", but this is something Anthropic tests for for safety evals on models. "If a model can independently conduct complex AI research tasks typically requiring human expertise—potentially significantly accelerating AI development in an unpredictable way—we require elevated security standards (potentially ASL-4 or higher standards)".

https://www.anthropic.com/news/announcing-our-updated-respon...

Re: GPT-5 is behind schedule

#887

Post by asdff » Mon, Dec 23, 2024, 7:39 PM UTC

palata wrote: post #594

> Altman told students in a talk at Stanford University that OpenAI could say with “a high degree of scientific certainty” that GPT-5 would be much smarter than the current model. I love that quote. "I can tell you with a high degree of scientific certainty that my product will be badass... for some definition of badass, please don't ask". I, too, could say with a high degree of scientific certainty that while puttin…

Imagine the ceo of intel going on stage and saying I can promise with a high degree of certainty the next cpu will be faster than the last. The fact that had to be said at all means the short sellers are making their money alright.

Re: GPT-5 is behind schedule

#888

Post by EVa5I7bHFq9mnYK » Mon, Dec 23, 2024, 7:42 PM UTC

josu wrote: post #864

EVa5I7bHFq9mnYK wrote: post #791

Earlier quoted context omitted.

In their attempt to block OpenAI, they block me. Many sites that were accessible just 2 years ago, require login/captchas/rectal exam now just to read the content.

> captchas I suspect that AIs are already more effective than humans at passing captchas.

That would be an example of AI providing real value that I would pay for.

Re: GPT-5 is behind schedule

#889

Post by 1vuio0pswjnm7 » Mon, Dec 23, 2024, 7:42 PM UTC

jaybna wrote: post #741

25% of the top 1000 websites are blocking OpenAI from crawling: https://originality.ai/ai-bot-blocking I am betting hundreds of thousands, rising to millions more little sites, will start blocking/gating this year. AI companies might license from big sources (you can see the blocking percentage went down), but they will be missing the long tail, where a lot of great novel training data lives. And then the big sites w…

Bill Gross:

https://twitter.com/Bill_Gross/status/1859999138836025808

https://pdl-iphone-cnbc-com.akamaized.net/VCPS/Y2024/M11D20/...

He appears to be criticising "AI" only to solicit support for his own company.

Re: GPT-5 is behind schedule

#890

Post by EVa5I7bHFq9mnYK » Mon, Dec 23, 2024, 7:44 PM UTC

ammanley wrote: post #841

EVa5I7bHFq9mnYK wrote: post #791

Earlier quoted context omitted.

In their attempt to block OpenAI, they block me. Many sites that were accessible just 2 years ago, require login/captchas/rectal exam now just to read the content.

Im looking forward to the life experience that is content I want to read badly enough to endure a rectal exam.

It's not that bad ...