This was my concern, with all the shenanigans MS would end up with more influence on open AI
it’s not clear to me where the problems start with that whole “philosophy”, only that there are indeed problems.
that’s my own swing and a miss at a causality joke. oh, well. i tried. have to start somewhere. oh god, not again.
i think the issue is people don’t know, so can’t say. the things his sister has said about him open the door for a wide range of behavior that a board might not like, and might not like to publicize either
I too want a job where if I’m ever fired, 95% of the workforce threatens to resign unless I get reinstated, pronto.
Having said that, WTF? This is some seriously weird stuff. Maybe I’ll stay in my current job after all.
Shit Clippy just got real.
https://twitter.com/edzitron/status/1727211248620839103
The story here is not that Sam Altman “won” anything back. This was a venture-backed ransacking of OpenAI - a judo move that made its board into a rogue’s gallery of growth at all costs executives and economists, and a publicity campaign for what will now be a $100bn+ valuation
sure thing…
OpenAI defines AGI as autonomous systems that surpass humans in most economically valuable tasks.
Oh, so that’s what true intelligence is about. I never realized it was an economics term. Thanks, capitalism!
It’s coming out that the boardroom controversy that got Altman fired may have centered around a potential emergent form of AGI from company research called Q* (or Q-star) that is able to solve math problems on its own in the same way humans do, but much better and faster. I can see how “free”-market fundies who also describe themselves as supreme rationalists – especially SV executives and VCs – would view this as an existential threat (at least to the humans they care most about).
i can’t wait till they learn how power plugs work.
“Off switches? Where we’re going (to the bank) we don’t need off switches!”
I remember when calculators could do arithmetic, but say factoring polynomials was something you had to do by hand. Now there are websites that do it much faster and more reliably for free. It feels kind of underwhelming as AI apocalypses go…unlike the glut of interpolated art and writing, I don’t think even mathematicians mind.
These are people who have no idea where the line between tool and general intelligence is, on top of having no idea where the line between general intelligence and deity is. As religions go, it makes far less sense than listening to a man who got lost in the desert and claims God told him not to eat beans, and it’s frustrating venture capital has made them matter.
I think the concern here is that the dingus can do these complex tasks without human participation or prompting. The worry isn’t entirely unfounded, in a paperclip maximising kind of way, even if one doesn’t consider this true AGI. However, as you implied, it’s more about MBAs and tech “visionaries” looking at it and saying:
Sure, assemble a vast pile of horseshit, and it will spontaneously generate ponies.
we dont know if this even slightly true; the last week was such an absurd shitshow, I dont believe anything right now what any of these people say.
there are no “plugs”…
/s
I still believe the main issue the board members had with Altman were the side-pocket ventures he was quietly involving himself in based on his OpenAI cachet. Still, I don’t put these Longtermist weirdos above making the business decision about Skynet-type gedankenexperiments too. The unusual corporate governance structure that allowed this debacle was grounded in TESCREAL thinking.
Ultimately, though, this will come down to everyone involved being afraid of their own corporate shadow.