Based on that article and the diagram of OpenAI’s structure, it looks like the board was focused on big-picture AI issues, Altman more on profit, so they fired him. And then other entities who are also more focused on profit, like Microsoft, freaked out. And the profit people won/are going to win. Is this roughly what’s happening?
Yea my understanding is that OpenAI was created with the specific goal of researching artificial intelligence in order to make sure it doesn’t destroy humanity. Then they saw how quickly it blew up and how much money they could make off it and decided (as humans do) ahh fuck that let’s just get rich.
I think you got it right.
It’s not really clear what the disagreement is because the board hasn’t really said anything about why they fired Altman. I’ve read some speculation that there were significant advances with GPT-4.5, and that the board is trying to lock down control of it for safety reasons. The situation is really bizarre.
The problem is they are trying to put the toothpaste back in the tube after the toothpaste factory exploded.
In an internal memo obtained by Axios on Saturday, OpenAI COO Brad Lightcap hinted at criticism of how the firing was handled and confirmed that Altman was not fired in response to “malfeasance or anything related to our financial, business, safety, or security/privacy practices.”
That paper looks reasonable for what it is, but I generally avoid consciousness debates because I do not have any strong intuitions or the inclination to acquire a well thought out postition. Something like:
“In a nutshell, our theory is that consciousness developed as a memory system that is used by our unconscious brain to help us flexibly and creatively imagine the future and plan accordingly,” explained corresponding author Andrew Budson, MD, professor of neurology. “What is completely new about this theory is that it suggests we don’t perceive the world, make decisions, or perform actions directly. Instead, we do all these things unconsciously and then—about half a second later—consciously remember doing them.”
Is probably generally correct, and it has some relation to Dennett’s Consciousness Explained, which, to my knowledge, is still the best book in the area. In this sense, I am a deflationist when it comes to consciousness, but I could be convinced otherwise quite easily.
Alarms go up for me when people start talking about “unconscious” processes or states, as that can mean many different types of things, some of them Freudian, and I really don’t like Freud, or at least anyone influenced by Freud. That said, I do generally believe that most inferences or judgments are “unconscious” and are interpreted and molded via subsequent language-influenced conscious reflection.
I have no idea about DID. I’m not sure what you describe as a “theory” of DID is a good approach. People are very promiscuous in proposing such “theories” for metal phenomena. Just look at “theories” of autism (eg, “autists lack a ‘theory of mind’”), or earlier theories for hysteria. Freudianism is basically a grab bag of building blocks for bullshit psychology theories. There are also “theories” of mental phenomena such as the “bicameral mind” conception of consciousness in Westworld. Scott Alexander, a very smart psychiatrist (who is sometimes completely wrong), often discuss such things in his Astral Codex Ten substack, as well as various “theories” for things like OCD. Some such theories may be correct, but I think a virtue of CBT, as I understand it, is that it is less theory and more results oriented.
Most of such theories seem to me akin to theories of alchemy, where you can describe some limited subset of phenomena with a “theory” that is more or less consistent and “feels” explanatory in its domain but does not generalize. I am at least sympathetic to eliminative materialist conceptions of mental theories, ie, that they are simply the wrong level of description.
OpenAI can decide to go all in on “AI Ethics”, but that will simply render them irrelevant. Like if Nobel had banned TNT production due to concerns about it’s potential harm we would still have nuclear weapons today. AI is an open arms race, and the arms race will continue even if a current leader drops out.
Also, I am not at all convinced that AI ethics is in any sense substantive beyond normal ethics (eg, don’t do race based facial recognition or profiling). Ethical treatment of AI issues will develop with a lag to actual AI developments, but as long as the primary supranational system of governance is the UN and Russia and China are on the security council, there will be no real limit on AI development.
This is insane. MSFT acquired a $90bn company for free! Also:
https://x.com/ilyasut/status/1726590052392956028?s=46&t=9xanL2tZoKj22erGoTuL4A
Wild stuff
They did invest 13 billion dollars but I get what you mean.
It definitely seems like the board was acting in accordance with their mission but was ham-handed and rash. And also, the horse has left the barn and possibly become sentient, so their attempts to halt/slow the destruction of society are just cute at best.
If not Sutskever then who was actually the mastermind behind the boardroom coup? Someone had to decide on this plan and sell the other board members on it. The Quora guy?
Maybe the next level AI is pitting them against each other. Well I guess it is, but I mean consciously.
What a monumental fuckup
Might the best example ever of totally self inflicted injury to a company. Only X competes.
Twitter would never admit their mistake and rehire someone.