521–530 of 596 posts
Earlier quoted context omitted.
Easier than that would just be downloading the torrent of all of Reddit through Sept 2023. https://academictorrents.com/details/89d24ff9d5fbc1efcdaf9d7...
The question is: if you then added all of Usenet before, say, 1992, would the effective intelligence of the trained LLM go up or down?
Earlier quoted context omitted.
Yes, potentially. The data has to come from somewhere, and all of the outlets that were used to train ChatGPT, stable diffusion, etc. have since been locked down. Any new company that Sam Altman makes in the AI space won't be competing just on merits of talent and product, they will also need to pay for and negotiate access to data. I'd actually expect this to get far worse going forward, now that other organizations…
Couldn’t he just partner with Microsoft and start with the Bing corpus?
Earlier quoted context omitted.
The point is that the data is easily accessible. If you wanted to get your hands on the data while simultaneously keeping them clean, contract with a Russian contracting company to give you a data dump. You don't need to know how they got it.
Subcontracting out your crimes isn’t going to fly in court.
The key question in my mind is not who is going to be on the new board, but whether Ilya Sutskever will stay if Altman comes back. I worry that OpenAI without Ilya is not going to produce groundbreaking innovations at the same pace. Hopefully Sam Altman and Ilya Sutskever can patch things up. That's more important than who they add or remove to the board.
This is a terrible take. Ilya staying is the only thing that is basically guaranteed. Everything else isn’t. In addition, no one is irreplaceable.
How is that guaranteed? If investors remove him from board of directors, he may get pissed off and quit, no?
> In addition, no one is irreplaceable.
In theory, maybe. In practice, it is not always easy. Nearly a year after ChatGPT came out Google hasn't been able to catch up. If it was easy to replace Ilya after he left Google, they would have caught up by now.
> With his firing from OpenAI, Altman quickly got the upper hand in terms of public messaging. The board didn’t use a communications or law firm in its dealings, people familiar with the board said, expecting that the OpenAI team would help them. But Altman had loyalty from investors and employees. https://www.wsj.com/tech/ai/openai-leadership-hangs-in-balan...
The loyalty was significant. The PR was meh
The board however has only made things ever more murky with their vague and fig-leaf like defenses. They appear utterly unprepared to deal with the aftermath of their own action, which they and only they knew was coming.
Many are speculating that Sam Altman could just move on and create another OpenAI 2.0 because he could easily attract talent and investors. What this misses is all the regulatory capture that he’s been campaigning for. All the platforms have now closed their gardens. Authors and artists are much more vigilant about copyright etc. So it’s now a totally different game compared to 3 years ago because the data is not jus…
I don't think getting training data is that hard still, the biggest platforms that locked down their APIs still use them for their mobile apps and can easily be reverse engineered to find keys or undocumented endpoints (or in the case of reddit, an entirely different internal API with less limits and a lot more info leaks...)
This seems to me like an example of how difficult it is to organize a company around a goal other then making money. As a non-profit, OpanAI was not supposed to be a profit maximizing enterprise. But how is a board supposed to opperate, and set objectives without a clear goal like profit maximization? Usually a boards represents the owners of the business and their interested. The OpenAI board does not represent the…
How is board membership controlled? For a profit company, you buy shares and elect the board, right? If they turn the company into something nobody wants, the shares lose value, and maybe someone picks up the shares cheap and turns it around. So there's a feedback cycle. The profit motive is almost incidental. But here do the current board members just appoint the next board members and grow/shrink the total number?
Earlier quoted context omitted.
That does not sound like the proper way to do an openAI 2.0. If Reddit ever hears that's how an AI company scraped them, they'll get sued for fun and profits.
It's essentially impossible to prove in court that training data was obtained or used improperly unless you go and tell on yourself. And even then it requires you to actually make someone with a lot of money mad, or to not have enough money yourself. Certainly microsoft would have already caught lots of flak for training their models on every github repo, instead they got a minor paddling from the public eye that wen…
Altman is more powerful than I thought.
I don't think so at all -- the board was simply far dumber than we thought. The board didn't plan or think this through at all . This isn't about Sam being powerful, just about him being a reasonable predictable cofounder Microsoft can work with. It's the rest of the board that shocked Microsoft with how unprofessional their actions were, that they can't be worked with. The very idea that they would fire Sam without…
Earlier quoted context omitted.
I think the hearts on Twitter was supposed to be a show of power on Sam's side. I wouldn't be surprised if that number is >20% of employees
That number is 91 employees out OAI's 700. Seems like a pretty reasonable number of people to support Sam's faction, but absolutely recoverable for OAI even if every single heart leaves.