Will California AI regulation bill SB 1047 become law this session?
➕
Plus
391
746k
Oct 2
51%
chance

California Senator Scott Weiner of SF has introduced the bill (https://twitter.com/Scott_Wiener/status/1755650108287578585, https://leginfo.legislature.ca.gov/faces/billNavClient.xhtml?bill_id=202320240SB1047).

This bill would regulate AI systems in various ways.

Will it become law by the time the legislative session ends?

This resolves YES if SB 1047 becomes law as a substantively similar bill. I wouldn't count it if it is gutted so much that it now essentially does nothing, but mostly I will let lesser changes stand.

This resolves NO if SB 1047 does not become law by the deadline and the session is over. If the session runs late the deadline will be extended.

Note that this applies only to SB 1047, if another similar bill is introduced and passes that would not count (it seemed too messy to worry about edge cases).

Get Ṁ1,000 play money
Sort by:

Any read through on this as to veto probability?

For those treating Pelosi's opposition as indication of Newsom's thinking, she might just be playing 4d chess once again.

https://sfstandard.com/2024/08/30/christine-pelosi-scott-wiener-shadow-campaign/

Some of Nancy Pelosi’s recent moves in San Francisco have been cast as an effort to derail Wiener’s quest for her House seat when she retires. Supposedly, the speaker emerita is clearing the way for her daughter, political strategist Christine Pelosi, to be her heir apparent.

...

Todd David, a longtime adviser to Wiener, said it’s incomprehensible that Pelosi’s pushback against the AI bill wasn’t a political calculation to gather support for her daughter’s possible bid for Congress.

“It’s explicitly a move to get Andreessen Horowitz and Google to fund an independent expenditure committee for Christine,” David said. “The motivation is, from where I sit, a political motivation. Not a policy motivation.”

@PlainBG Pelosi kinda looking like a snack ngl

Is it hovering around 50/50 because Zvi said it was a tossup, or was it already hovering there before that?

@JS_81 Newsom is undecided 50/50 because Zvi said it was a tossup

@HenriThunberg Honestly I wish Newsom obeyed the whims of Zvi. California would be a better place.

The description says "Will it become law by the time the legislative session ends?"
Ballotpedia says:

> In 2024, the California State Legislature was scheduled to convene on January 3 and adjourn on August 31.

So since the bill hasn't become law yet as of today, I would assume this should resolve No. But the market is still at 50%. Am I misunderstanding what the cutoff date is for "by the time the legislative session ends"?

@yetforever I have ruled that this will resolve based on Newsom and his veto or signing of the bill.

@ZviMowshowitz Thanks! I have a follow-up question then:

I believe the legislature can potentially override a veto. If Newsom vetoes the bill, am I understanding correctly that this market immediately resolves to No, regardless of how the legislature chooses to respond?

@yetforever I have been assured that won't happen here and I won't wait for it.

@ZviMowshowitz You have been assured the legislature won’t override the veto? How have you been assured of that? Lol

(Although I understand the not-waiting-for-it part, I have no issue with that)

@benshindel Since 1980, there have been over 8,000 vetoes by California governors, and zero have been overridden by the legislature.

https://sor.senate.ca.gov/sites/sor.senate.ca.gov/files/3258%20SOR%20governors%20veto%20report%202022.pdf

@benshindel There is no time to do it, also there aren't enough YES votes.

Is there a market for when it passes? Like a date interval? It's been going on for a while and I want to know when I might come back to this.

It occurs to me that this will plausibly not be settled by a tug of war of which side can accumulate the most support to sshow Newsom, but rather by a question of 1) overall, which side is it least risky for Newsome to support? and 2) can that side accumulate the minimum satisfactory support where Newsom can support it while having deniability if things go wrong later?

(This updates me toward veto being less likely)

Reasons why I think p(veto) < 50%: Newsom seems fairly intelligent to me, Weiner seems very intelligent as well and seems like he really did his homework, Newsom will be convinced that the bill won't really harm the AI industry (Anthropic's statement and maybe Elon's will help), Newsom will be able to communicate honestly that the bill has been weakened and narrowed in scope to the people who opposed it before, and I think opposing the bill and then some even minor catastrophe happening will have a higher chance to blow back on Newsom if he personally vetoed a safety bill that was passed by all other relevant entities, and that will factor into his decision.

bought Ṁ500 YES

Poll: 7 in 10 Californians Support SB1047, Will Blame Governor Newsom for AI-Enabled Catastrophe if He Vetoes - AI Policy Institute (theaipi.org)

SB1047 stifling the innovation feels more fuzzy even if real

its easier to blame him for any shit happening from AI (even outside the bill's scope) than for hindering AI progress, he can't know which will happen but still can manage his personal risks

but I'm silver

@DanielPolak This is basically the way I'm thinking about this question. Gavin Newsom's utility function is "maximize the chances that Gavin Newsom is someday elected President" and while big tech money can help with that, being seen as too friendly with big companies (especially one that is fully on team Trump like a16z) is I think pretty dangerous in a Democratic primary. "Gavin Newsom sides with MAGA big tech companies to kill AI regulation" is a pretty difficult headline to have in 2028.

@SemioticRivalry "Gavin Newsom stifled economic growth (with an unconstitutionally vague law? Perhaps?) against the recommendation of experts in the field and senior members of his party" is also a really difficult headline.

I could be wrong, but I think there's a decent chance he vetoes still.

@Najawin I think there's a decent chance he vetoes it, maybe 35%. But it's a lot harder to get headlines about overregulation than underregulation, especially in a democratic primary. Being too tough on big tech is not going to be a problem for primary voters. I think that AI is a bigger issue in 4 years and the Democratic party is fairly likely to lean more anti-AI due to jobs, actors, artists, and skepticism of corporations. This veto could be a real liability if he's seen as beholden to the industry.

@SemioticRivalry The other aspects to this are what makes it a harsher line of attack imo. It could show an unwillingness to take expert advice or to listen to others in the party who have real concerns.

i think it's naive to cite popular support as a meaningful indicator for likelihood a bill gets passed

@DanielPolak Newsom doesn’t have to appeal to cali voters for presidential aspirations, but the whole country.

A large amount of AI tech leaders find the safety argument silly at this point, and a bill like this threatens to undermine innovation on the way to AGI, which has much higher potential at helping humanity achieve post-scarcity than destroy us. People watch too much sci fi.

Newsom will have to answer for why he caused several large cali tech companies to leave the state entirely because of incredibly speculative risk assessment more based on the plot of Terminator 2 than actual real risks. If AGI is invented and deems humanity a threat, then human government is unlikely to be relevant at all anyway.

Wiener remarked that regulators often are playing catch up with risks and in this case should pre empt it, but the reason that is the case is because the risks of a new technology take time to understand. This legislation is no better than the police in minority report.

@JamieCrom I don't think that's naive? I'd guess that popular support strongly correlates with the likelihood of a bill getting passed, lol?

@Cole1 "Newsom will have to answer for why he caused several large cali tech companies to leave the state entirely"

This won't happen, though, because -- at least to my understanding -- nothing in the bill disadvantages companies based in California compared to companies based elsewhere. The law applies equally to all companies around the world that allow Californians to access their services. I think Newsom is probably smart enough to realize this.

@benshindel they data shows that popular support has 0 correlation with likelihood of passing

@EricNeyman moreover, no tech company would leave California over this law. It's really not that burdensome. Of all the regulations in California and the US affecting the AI industry, this one wouldn't rank highly. I think copyright law is like 100x more concerning to them.

bought Ṁ50 YES at 50%