Terrible news...
#SB1047 has been vetoed.
1 messages · Page 1 of 1 (latest)
As noted elsewhere, very disappointing.
The best thing to work on this is to hope that we can coordinate with our coalition and expand on it
He did say it was a good start, but too focused on current industry leaders, leaving smaller models to grow and make the same mistakes.
It was pretty much a lie
Ah. Oh well
Notably everyone in the CA legislature seemed to be better on this than he was.
Anyway, we now have to lobby hard on the federal side.
Sad but unsurprising. We're in a race. Every action that could be seen as potentially slowing down AI research for some country or state will face the same criticism. The same selfish reasons will be used again and again.
This is why we've been saying for over a year now that we need international regulation. We will not get sufficient safety measures on a national or state level. The incentives of racing ahead are too powerful. The game theory only works out if we all cooperate. Let's get out of this prisoners dilemma.
So focus on the UN?
Not as a target - we still have to convince our national leaders. We need them to initialize treaty negotations
I’d still imagine that a convinced person inside the UN could do a fair bit of mobilizing nations
The upcoming SF summit in november, and finding out who is attending - that should probably be our main focus now
#1288842991661744169
"By focusing only on the most expensive and large-scale models, SB 1047 establishes a regulatory framework that could give the public a false sense of security. Smaller, specialized models may emerge as equally or even more dangerous".
What the fuck Gavin
"This bill only bans the largest bombs. Small bombs may be just as dangerous. So let's not ban large bombs"
Newsom: we also need to regulate small models and companies
Pelosi: thanks for not regulating small models and companies
???
This part really sticks out to me.
I honestly read that as “California is afraid of losing their largest donors by holding them accountable, so let’s just not.”
How else is anyone supposed to interpret this?
The Nirvana fallacy when it comes to AI regulation is still so strong. Gavin is fine with it as long as it's a perfect piece of legislation which covers all possible dangers while having zero tradeoffs. And there's a unicorn which farts rainbows.