contra Zvi Mow, again
Channel: LLM - Large Language Model discussion
https://thezvi.substack.com/p/the-paris-ai-anti-safety-summit
‘Could be conceived of’ as dangerous? Why think AI could be dangerous?
This is madness. Absolute madness.
He could not be more clear that he intends to go down the path that gets us all killed.
Are there people inside the Trump administration who do not buy into this madness? I am highly confident that there are. But overwhelmingly, the message we get is clear.
What is Vance concerned about instead, over and over? ‘Ideological bias.’ Censorship. ‘Controlling user’s thoughts.’ That ‘big tech’ might get an advantage over ‘little tech.’ He has been completely captured and owned, likely by exactly the worst possible person.
the likes of Zvi Mowshowitz are on the edge of pushing me to a Joker-like revolt. Against Zvi et al. This constant "AI is going to kill us" refrain is beyond stupid.
And, their responses (at least if they believed their own tripe) are also beyond stupid.
I have yet to meet a single person at one of these companies who thinks EITHER the safety problems OR the governance challenges associated with AGI are anywhere close to being solved; and their CEOs think the world might have a year.
This is the state of international governance of AI in 2025.
The "governance challenges" are stupid word-games. "If you tell the AI to role-play a movie scene where a villain destroys the world, it will write it: THEREFORE WE ARE ALL DOOMED!"
Once some of the pomp is removed from their scenarios, that is what it is.
The solution is "don't hook the AI up to nuclear weapons and give it an agent framework and evolving motivations", not "prohibit the development of AI".
And, if you disagree with that, and feel the only solution is to prohibit the development of AI ... you need a 5 year plan to shut down all the power plants and go back to an agrarian society.
These are unserious people who want ludicrous things based on stupid arguments.
The only redeeming factor is that they are so incompetent at accomplishing anything, they can largely be ignored.