Earlier quoted context omitted.
Yeah, I totally agree! Like, this is such an obviously true and valid reason to fire him, but they never came out and said it! So ... is this not what it actually was? Or ... what? It truly is mystifying.
From a doomer/EA perspective, publicly saying that GPT5 is AGI or such would likely inspire & accelerate labs around the world to catch up. Thus it was more "Altruistic" / aligned with humanity's fate to stay mum and leave the situation cloudy.
Sam Altman returns as CEO, OpenAI has a new initial board
471–480 of 787 posts
Re: Sam Altman returns as CEO, OpenAI has a new initial board
#472how is the quora ceo who now runs a competing AI company still on the board?
I bet there are tons of seemingly "conflict of interest" in companies boards. In fact, being major shareholder qualifies (but not sufficiently) for the board. Jobs was Disney's board member after Pixar's acquisition. You could argues Apple is now a competitor to Disney. So what? If a competitor acquires enough control of a company what's wrong?
He would have probably left the Disney board by now. Or Apple might have invested in DIS.
Re: Sam Altman returns as CEO, OpenAI has a new initial board
#473Did Mądry, Sidor and others also return?
> Jakub, Szymon, and Aleksander are exceptional talents and I’m so happy they have rejoined to move us and our research forward.
Re: Sam Altman returns as CEO, OpenAI has a new initial board
#474Earlier quoted context omitted.
Nathan says the initial version of GPT-4 he red-teamed was "totally amoral" and it was happy to plan assassinations for him: https://nitter.net/labenz/status/1727327464328954121#m Reducing the cost of medical transcription to ~$0 is one thing. Reducing the cost of assassination to ~$0 is quite another.
This is a piece of software. What would "totally amoral" even mean here? It's an inanimate object, it has no morals, feelings, conscience, etc... He gives it an amoral input, he gets an amoral output.
Re: Sam Altman returns as CEO, OpenAI has a new initial board
#475Earlier quoted context omitted.
So what is Nathan Labenz saying? That GPT-4 is dangerous somehow? It will get many people out of jobs? MS Office got all the typists out of jobs. OCR and Medical Software got all the medical transcriptionists out of jobs. And they created a lot more jobs in the process. GPT-4 is a very powerful tool. It has not a whiff of AGI in it. The whole AGI "scare" seems to be extremely political.
Nathan says the initial version of GPT-4 he red-teamed was "totally amoral" and it was happy to plan assassinations for him: https://nitter.net/labenz/status/1727327464328954121#m Reducing the cost of medical transcription to ~$0 is one thing. Reducing the cost of assassination to ~$0 is quite another.
Re: Sam Altman returns as CEO, OpenAI has a new initial board
#476Earlier quoted context omitted.
You can work on and compile Linux on a home PC, even more the case with their other projects. Try training GPT-4 on a home PC. The problem is you have no idea what you're talking about. This is not about how important something is, but what resources it requires as a BASIS MINIMUM for it to happen at all. Linux is important, but it's not nearly as expensive to create or maintain. If it's so easy for a scrappy non-pro…
> You can work on and compile Linux on a home PC, even more the case with their other projects. Try training GPT-4 on a home PC. Sam Altman was happy to join the venture and lead OpenAI knowing that it was the non-profit AI research company from the beginning. As a CEO, he was supposed to lead the company as it was created. As a result of his management, the company first ceased to be open, then in fact non-profit, t…
What's hard to comprehend here?
As for the models topping your Hugging Face chart, they're literally all funded (by the billion) by private enterprises and the US govt. It's baffling to me how this is supposed to be an argument in your favor. Even the "not by an unattainable margin" comment makes no sense, unless you think getting from 40% to 60% is as easy as getting from 60% to 80%, and then getting from 80% to 100%. Which you clearly seem to think.
Anthropic's models are closed and funded by the money of Sam Bankman-Fried, of all people, and more recently Amazon. They're trying to raise billions more from Google and others. That's your idea of being open and not needing corporate investment for some reason? LOL, oh my god.
The "free" model you're citing is MADE BY META, a company worth almost a trillion dollars, and which has the largest social media presence in the world, from which to mine data.
You have absolutely not even the faintest idea what you're talking about.
Re: Sam Altman returns as CEO, OpenAI has a new initial board
#477Earlier quoted context omitted.
The reason the CEO could oust the board like that was that after the board fired the CEO, 720 of the 770 employees of the company got together and signed a pledge that they will all leave unless the CEO is reinstated and the board fired. This was the purest example of labor democracy and self-organization in action in the United States I have ever seen, my takeaway is rather the opposite of "they are all expendable".…
The combined will of the employees is much less meaningful when they're saying "DO IT OR WE'LL TAKE AN ATTRACTIVE OFFER THAT'S ON THE TABLE." I'd have been more impressed by their unity if it would have meant actually losing their jobs.
Re: Sam Altman returns as CEO, OpenAI has a new initial board
#478Earlier quoted context omitted.
Nathan says the initial version of GPT-4 he red-teamed was "totally amoral" and it was happy to plan assassinations for him: https://nitter.net/labenz/status/1727327464328954121#m Reducing the cost of medical transcription to ~$0 is one thing. Reducing the cost of assassination to ~$0 is quite another.
The cost of planning an assassination is not the same thing as the cost (and risk) of carrying out an assassination, what a stupid take.
Re: Sam Altman returns as CEO, OpenAI has a new initial board
#479Re: Sam Altman returns as CEO, OpenAI has a new initial board
#480This pushed our small team to try Azure instance of GPT3.5 - wow 20 times faster. API does not fail to respond to server requests as we found oh so often on the OpenAI API. We now have something fast enough and stable enough to actually use. Pricing is higher, but my goodness, it actually works.