OpenAI is not far ahead from alternatives... it doesn't matter much what the outcome will be.
Unless they develop an agent that can recursively self-improve to the point of an intelligence explosion!
OpenAI and Elon Musk
301–310 of 1001 posts
Re: OpenAI and Elon Musk
#302Earlier quoted context omitted.
In my eyes this is a straw argument. "[T]otally OK not to share the science." I think the reasonable average person would disagree with that. And, it would go against certain goal & financial transparency principles that the IRS demands to bestow the 501(c)3 designation. (e.g. see here https://www.citizen.org/article/letter-to-california-attorne... )
Ilya's justification for that argument was to link to a SlateStarCodex blog post. We are doomed...
Where did you get that it was sent by Ilya? The to: field is redacted.
Re: OpenAI and Elon Musk
#303Earlier quoted context omitted.
> To some extent, they may be right that open sourcing AGI would lead to too much danger. They claimed that about GPT-2 and used the claim to delay its release.
They claimed that GPT-2 was probably not dangerous but they wanted to establish a culture of delaying possibly-dangerous releases early. Which, good on them!
Do you really think it is a coincidence that they started closing down around the time they went for-profit?
Re: OpenAI and Elon Musk
#304Earlier quoted context omitted.
Do you trust a for profit corporation with the codification?
Call me a capitalist, but I trust several of them competing with each other under the enforcement of laws that impose consequences on them if they produce and distribute content that violates said laws.
Wait but who codifies the ethics in that setup? Wouldn’t it still be, at best, an agreement among the big players?
Re: OpenAI and Elon Musk
#305Earlier quoted context omitted.
> To some extent, they may be right that open sourcing AGI would lead to too much danger. That's clearly self-serving claptrap. It's a leveraging of a false depiction of what AGI will look like (no ones really knows, but it's going to be scary and out of control!) with so much gatekeeping and subsequent cash they can hardly stop salivating. No strong AI (there is no evidence AGI is even possible) is not going to be a…
> there is no evidence AGI is even possible Reading this is like hearing "there is no evidence that heavier-than-air flight is even possible" being spoken, by a bird . If 8 billion naturally-occuring intelligences don't qualify as evidence that AGI is possible, then is there anything that can qualify as evidence of anything else being possible?
we also cannot build most birds
Re: OpenAI and Elon Musk
#306Reading the messages: This is a marketing step of course, no sane lawyer would agree to this. And that is because, I don't think they show what they think and want to show. That at some point Elon had an opinion that a lot of money is needed or that OpenAI maybe had no future? That does not change the duty or obligations of a non-profit to the mission. Also, it is clear some important information has been blacked out…
And what would happen? 2 minutes after being dissolved, ClosedAI is founded and hires all the employees.
Re: OpenAI and Elon Musk
#307Their 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're probably right that building AGI will require a ton of computational power, and that it will be very expensive. Why? This makes it seem like computers are way less efficient than humans. Maybe I'm naive on the matter, but I think it's possible for computers to match or surpass human efficiency.
Perhaps the finalized AGI will be more efficient than a human brain. But training the AGI is not like running a human, it's like speed running evolution from cells to humans. The natural world stumbled on NGI in a few billion years. We are trying to do it in decades - it would not be surprising that it's going to take huge power.
Re: OpenAI and Elon Musk
#308Earlier quoted context omitted.
Counter-counterpoint: absolutely nobody who has unguardrailed Stable Diffusion installed at home for private use has ever asked for more guardrails. I'm just saying. :) Guardrails nowadays don't really focus on dangers (it's hard to see how an image generator could produce dangers!) so much as enforcing public societal norms.
Just because something is not dangerous to the user doesn’t mean it can’t be dangerous for others when someone is wielding it maliciously
What kind of damage can you do with a current day llm? I’m guessing targeted scams or something? They aren’t even good hackers yet.
Re: OpenAI and Elon Musk
#309> For example, Albania is using OpenAI’s tools to accelerate its EU accession by as much as 5.5 years What an insane thing to say so matter of factly. This is like a character in an airport bookstore political thriller who is poorly written to be smart.
Why is that such an insane thing for them to say?
Re: OpenAI and Elon Musk
#310Earlier quoted context omitted.
> More controls make LLMs more trustworthy for many people, not less. The Snafu at Goog suggests a need for improved controls, not 0 controls. To whom? And, as hard as this is to test, how sincerely? > The American culture wars are not global. (They have their own culture wars). Do people from places with different culture wars trust these American-culture-war-blinkered LLMs more or less than Americans do?
- To me, the teams I work with and everyone handling content moderation. / Rant / Oh God please let these things be bottle necked. The job was already absurd, LLMs and GenAI are going to be just frikking amazing to deal with. Spam and manipulative marketing has already evolved - and thats with bounded LLMs. There are comments that look innocuous, well written, but the entire purpose is to low key get someone to do a…
I hope you don't delete it! I enjoyed reading it. It pleased my confirmation bias, anyways.
Your comment might help someone notice patterns that they've been glancing over....
I liked it up until the T&S part. My eyes glazed over the rest since I didn't know what T&S means. But that's just me.