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

OpenAI and Elon Musk

openai.com

Post reply on HN
Print view View original

441–450 of 1001 posts

Re: OpenAI and Elon Musk

#441

Post by leereeves »

sigmoid10 wrote: post #408
leereeves wrote: post #388

Earlier quoted context omitted.

More compute also requires more data - scaling equally with model size, according to the Chinchilla paper. How much more data is available that hasn't already been swept up by AI companies? And will that data continue to be available as laws change to protect copyright holders from AI companies?

It's not just the volume of original data that matters here. From empirics we know performance scales roughly like (model parameters)*(training data)*(epochs). If you increase any one of those, you can be certain to improve your model. In the short term, training data volume and quality has given a lot of improvements (especially recently), but in the long run it was always model size and total time spent training th…

In smaller models, not having enough training data for the model size leads to overfitting. The model predicts the training data better than ever, but generalizes poorly and performs worse on new inputs.

Is there any reason to think the same thing wouldn't happen in billion parameter LLMs?

Re: OpenAI and Elon Musk

#442

Post by BlueTemplar »

Atotalnoob wrote: post #203
chatmasta wrote: post #185

Earlier quoted context omitted.

Ilya's justification for that argument was to link to a SlateStarCodex blog post. We are doomed...

What’s wrong with slatestarcodex? I’ve never heard of them before

Speaking of, I have been wondering how much LLM breakthroughs were helped by Karl Friston ??

https://slatestarcodex.com/2019/03/20/translating-predictive...

Re: OpenAI and Elon Musk

#443

Post by frabcus »

ascorbic wrote: post #404
Barrin92 wrote: post #353

Earlier quoted context omitted.

because it's a completely made up claim with a random number thrown in. How do you end up with "5.5" years for a process that doesn't have a timetable, number of estimated emails required for EU accession divided by number of emails ChatGPT can generate per day?

I was pretty skeptical about this too, but looking into it there is some basis in fact. They're using it to help translate and integrate the "Acquis communautaire" - the huge body of EU laws and regulations that need to be enshrined in national laws of candidate countries. This is one of the most time-consuming parts of the process, and usually takes many years. Leaving aside how risky this is (presumably they will h…

An article about this with some more details:

https://www.euractiv.com/section/politics/news/albania-to-sp...

Re: OpenAI and Elon Musk

#444

Post by hef19898 »

ben_w wrote: post #423
enriquec wrote: post #208

Earlier quoted context omitted.

[flagged]

It's way more complex than that. The Communist Manifesto was published during the Great Famine in Ireland, and that famine was much worse than it needed to be because the UK government didn't intervene. And a big part of the growth in capitalist societies is related to corporate structures, which are small scale collectives, with bosses who make decisions for all. When the US civil war happened, that resulted in the…

When you look at how by the book Communism, not necessarily the one Marx wrote about, runs their economies, you start to see a lot of parallels to how companies run their businesses. The mistake is to apply those priciples at too high a level, as it takes away some of the individual decisions and incentives, and not everything can be managed centrally, especially the customer demand side.

Communist industry and economy worked reasonably well for the stuff were the state is the natural customer in any society: defence. Everything else, not so much.

Re: OpenAI and Elon Musk

#445

Post by jbc1 »

carlossouza wrote: post #108
magnio wrote: post #6

> Elon understood the mission did not imply open-sourcing AGI. As Ilya told Elon: “As we get closer to building AI, it will make sense to start being less open. The Open in openAI means that everyone should benefit from the fruits of AI after its built, but it's totally OK to not share the science...”, to which Elon replied: “Yup”. Wow the most open OpenAI has ever been is when someone sue them. On the other hand, th…

> The Open in openAI means that everyone should benefit from the fruits of AI after its built, but it's totally OK to not share the science... Based on this line of reasoning, ANY company that builds any given technology and intends to share (sell) it to the world, but not divulge how it was done, can call itself OpenWhatever. They are clearly saying that the word “Open” in their name means nothing.

Any company could do that. If it would make sense for them to do that would depend on the market they were entering though. At the time OpenAI came about companies weren't sharing (selling) AI to the world. Doing so was a point of differentiation. There's Google over there hoarding all of their AI for themselves. Here's us over here providing APIs and free chat interfaces to the general public.

So sure the name means nothing now in a market shaped by OpenAI, where everyone offers APIs and has chat interfaces. It doesn't mean it meant nothing when they picked it or that they abandoned the meaning. The landscape just changed.

Re: OpenAI and Elon Musk

#446

Post by BigParm »

elif wrote: post #56
Reubend wrote: post #13

Their evidence does a nice job of making Musk seem duplicitous, but it doesn't really refute any of his core assertions: they still have the appearance of abandoning their core mission to focus more on profits, even if they've elaborated a decent justification of why that's necessary to do. Or to put it more simply: here they explain why they had to betray their core mission. But they don't refute that they did betra…

>they may be right that open sourcing AGI would lead to too much danger. I think this part is proving itself to be an understandable but false perspective. The hazard we are experiencing with LLM right now is not how freely accessible and powerfully truthy it's content is, but it is precisely the controls upon it which are trying to be injected by the large model operators which are generating mistrust and a poor und…

I think that’s missing the main point which is we don’t want the ayatollah for example weaponizing strong AI products.

Re: OpenAI and Elon Musk

#447

Post by dkarras »

TheAlchemist wrote: post #285
dkjaudyeqooe wrote: post #146

Earlier quoted context omitted.

The difference is OpenAI had a reputation to protect, Musk can't sink any lower at this point, and his stans will persist. The fact that they slinging mud at each other just proves the mutual criticism though, and provides an edifying spectacle for the rest of us.

> Musk can't sink any lower at this point, and his stans will persist. You sure about that ? For now they are largely looking away (and boy it's hard !) from his 'far right' adventures. But it was just reported that he met with Trump. And it's pretty clear that if Trump is elected and does cancel EV subsides as he says he will do, Tesla is dead and he knows it. So now, we have both of those guys, having something eac…

I mean his standing is based on what he is capable of, not what he does given the circumstance. We know he would side with Trump if it benefits him. I don't need to see the scenario play out for me to judge him for it.

Re: OpenAI and Elon Musk

#448

Post by finnjohnsen2 »

lagt_t wrote: post #105
Reubend wrote: post #13

Their evidence does a nice job of making Musk seem duplicitous, but it doesn't really refute any of his core assertions: they still have the appearance of abandoning their core mission to focus more on profits, even if they've elaborated a decent justification of why that's necessary to do. Or to put it more simply: here they explain why they had to betray their core mission. But they don't refute that they did betra…

Everytime they say LLMs are the path to AGI, I cringe a little.

Yea the idea that the computers can truly think by mimicking our language really well doesn't make sense.

But the algorithms are black box to me, so maybe there is some kind of launch pad to AGI within it

Re: OpenAI and Elon Musk

#449

Post by hef19898 »

eightman wrote: post #374
TaylorAlexander wrote: post #275

Earlier quoted context omitted.

The biggest real problem I’m experiencing right now isn’t controls on the AI, it’s weird spam emails that bypass spam filters because they look real enough, but are just cold email marketing bullshit: https://x.com/tlalexander/status/1765122572067434857 These systems are making the internet a worse place to be.

The use case for AI was, is and always will be spam.

You forgot the initial use case for the internet: porn.

Re: OpenAI and Elon Musk

#450

Post by hef19898 »

Jensson wrote: post #438
bergen wrote: post #272

Earlier quoted context omitted.

Do you truely believe google, OpenAI or Ford are the product of a single person?

Everyone contributed for the sake of themselves and not for the greater good of the company, that is what individualism is.

Ypu onoy have collegues like that? Sounds quite toxic.
Post reply on HN