Gavin Newsom vetoes SB 1047
gov.ca.gov
gov.ca.gov
https://news.ycombinator.com/item?id=41690454 remains pinned to the top of the thread, so people have the opportunity to read both.
(Actually we usually prefer the best third-party article to press releases, but nothing's perfectly consistent.)
I'm repeatedly seeing A.T. links posted which read "you have been blocked" or similar. Sometimes those resolve later, sometimes not.
HN's policy is that paywalls are permissible where workarounds exist. WSJ is getting close to disabling those workarounds.
The NYTimes similarly tightened its paywall policy a few years ago. A consequence was that its prevalence on the HN front page fell to ~25% of its prior value, with no change in HN policies (as reported by dang), just member voting patterns.
Given the difficulty in encouraging people to read articles before posting shallow-take comments, this is a significant problem for HN, and the increased reliance of media sites on paywalls is taking its toll on general discussion.
There are literally hundreds of news sites, and many thousands if individual sites, submitted to HN and making the front page annually. It would cost a fortune, not merely a small one, to subscribe to all of these.
No one has the time to read all of them, so it doesn't really matter if it would also be unaffordable.
HN takes pride in being both a single community and discussing a wide range of sources. Wider adoption of subscriber paywalls online would be inimical to both aspects.
It was basically a torpedo against open models. Market leaders like OpenAI and Anthropic weren't really worried about it, or about open models in general. Its supporters were the also-rans like Musk [1] trying to empty out the bottom of the pack, as well as those who are against any AI they cannot control, such as antagonists of the West and wary copyright holders.
[1] https://techcrunch.com/2024/08/26/elon-musk-unexpectedly-off...
Actually, wait, if Grok is losing to GPT, why would Musk care about Llama more than Altman? Llama hurts his competitor...
Should we crown one the king?
Or perhaps it is better to let them compete?
Perhaps advanced AI capability will motivate advanced AI safety capability?
The flop targets in that bill would be like saying “640KB of memory is all we will ever need” and outlawing anything more.
Imagine what other countries would have done to us if we allowed a monopoly like that on memory in 1980.
One of those is $100MM in training costs.
The other is measured in FLOPs but is already larger than GPT-4, so the “think of the small guys!” argument doesn’t make much sense.
How do people on a computer technology forum ignore the 10,000x improvement in computers over 30 years due to advances in computer technology?
I could understand why politicians don’t get it.
I should think that computer systems companies would be up in arms over SB 1047 in the same way they would be if the government was thinking of putting a cap on hard drives bigger than 1 TB.
It puts a cap on flops. Isn’t the biggest company in the world in the business of selling flops?
Inversely to your sarcastic “understanding” about politicians’ stupidity, I can’t understand how tech people seem incapable or unwilling to actually read the legislation they have such strong opinions about.
We periodically raise flop limits in export control law. The intention is still to limit China and Iran.
Would any computer industry accept a government mandated limit on perf?
Should NVIDIA accept a limit on flops?
Should Pure accept a limit on TBs?
Should Samsung accept a limit on HBM bandwidth?
Should Arista accept a limit on link bandwidth?
I don’t think that there is enough awareness that scaling laws tie intelligence to these HW metrics. Enforcing a cap on intelligence is the same thing as a cap on these metrics.
https://en.m.wikipedia.org/wiki/Neural_scaling_law
Has this legislation really thought through the implications of capping technology metrics, especially in a state where most of the GDP is driven by these metrics?
Clearly I’m biased because I am working on advancing these metrics. I’m doing it because I believe in the power of computing technology to improve the world (smartphones, self driving, automating data entry, biotech, scientific discovery, space, security, defense, etc, etc) as it has done historically. I also believe in the spirit of inventors and entrepreneurs to contribute and be rewarded for these advancements.
I would like to understand the biases of the supporters of this bill beyond a power grab by early movers.
Export control flop limits are designed to limit the access of technology to US allies.
I think it would be informative if the group of people trying to limit access of AI technology to themselves was brought into the light.
Who are they? Why do they think the people of the US and of CA should grant that power to them?
Your delusions and lack of nuance shown in this very thread are exactly why people want to regulate this field.
If developers of nuclear technology were making similar arguments, I bet they’d have attracted even more aggressive regulatory attention. Justifiably, too, since people who speak this way can't possibly be trusted to de-risk their own behavior effectively.
There is a reason why we report time (speedup) in spec instead of $$
The price you pay depends on who you are and who is giving it to you.
With the other threshold, it creates a disincentive for models like llama-405B+, in effect enshrining an even wider gap between open and closed.
And even if it were, if said guy has such amount of compute, then it's time to use some of it to describe the model's safety profile.
If it makes sense for Meta to release models, it would have made sense even with the requirement. (After all the whole point of the proposed regulation is to get some better sense of those closed models.)
A much better objection to static FLOP thresholds is that as data quality and algorithms increase, you can do a lot more with fewer FLOPs / parameters.
But let’s be clear about these objections - they are saying that FLOP thresholds are going to miss some harms, not that they are too strict.
The rest is arguing about exactly where the FLOP thresholds should be. (And of course these limits can be revised as we learn more.)
Like if I’m a startup reliant on open-source models I realize I don’t need liability and extra safety precautions but I didn’t hear any guarantees that this wouldn't turn off Meta from releasing their models to me if my business was in California?
I never heard any clarifications from the Pro groups about that
There’s no new disincentive to open sourcing models produced by this bill, AFAICT.
The thing that large companies care the most about in the legal realm is certainty. They're obviously going to be a big target of lawsuits regardless, so they want to know that legislation is clear as to the ways they can act - their biggest fear is that you get a good "emotional sob story" in front of a court with a sympathetic jury. It sounded like this legislation was so vague that it would attract a hoard of lawyers looking for a way they can argue these big companies didn't take "reasonable" care.
There are a number of related threads linked, but I’ll personally highlight Jeremy Howard’s open letter as IMHO the best-argued case against SB 1047.
Who are these wide range of >$100mm open source models he's thinking of? And who are the impacted small businesses that would be scared to train them (at a cost of >$100mm) without paying for legal counsel?
[1] https://en.wikipedia.org/wiki/Defense_of_Marriage_Act
[2] https://en.wikipedia.org/wiki/Personal_Responsibility_and_Wo...
[3] https://en.wikipedia.org/wiki/Illegal_Immigration_Reform_and...
[4] https://en.wikipedia.org/wiki/Dietary_Supplement_Health_and_...
"Thanks to fair use, you have a legal right to use copyrighted material without permission or payment. But thanks to Section 1201, you do not have the right to break any digital locks that might prevent you from engaging in that fair use. And this, in turn, has had a host of unintended consequences, such as impeding the right to repair."
https://www.eff.org/deeplinks/2020/07/what-really-does-and-d...
that's the worst
I’d be careful with that cognitive bias, because obviously companies dumping poison into water sources are going to be opposed to laws that would prohibit them from dumping poison into water sources.
Always consider the broader narrative in addition to the specific narratives of the players involved. Personally, I’m on the side of the fence that’s grumpy Newsom vetoed it, because it stymies the larger discussion about regulations on AI in general (not just LLMs) in the classic trap of “any law that isn’t absolutely perfect and addresses all known and unknown problems is automatically bad” often used to kill desperately needed reforms or regulations, regardless of industry. Instead of being able to build on the momentum of passed legislation and improve on it elsewhere, we now have to deal with the giant cudgel from the industry and its supporters of “even CA vetoed it so why are you still fighting against it?”
Generally speaking, it’s in ours’ and the company’s best interests to keep as little data as possible for two big reasons: legal discovery and cost. Unless we’re explicitly required to retain historical records, it’s a legal and fiscal risk to keep excess data around.
That said, there are situations where your input is captured and stored regardless of whether it’s sent. As you said, whether it does or not is often a simple search away.
When framed like that: why be opposed, it hurts your competition? The answer is something like: it shrinks the pie or reduces the growth rate, and that's bad (for them and others)
The economics of this bill aren't clear to me (how large of a fine would Google/Microsoft pay in expectation within the next ten years?), but they maybe also aren't clear to Google/Microsoft (and that alone could be a reason to oppose)
Many of the ai safety crowd were very supportive, and I would recommend reading Zvi's writing on it if you want their take
He vetoed it in part because the threshold it applies to at all are well-beyond any current models, and he wants something that will impose greater restrictions on more and much smaller/lower-training-compute models that this would have left alone entirely.
> Market leaders like OpenAI and Anthropic weren't really worried about it, or about open models in general.
OpenAI (along with Google and Meta) led the institutional opposition to the bill, Anthropic was a major advocate for it.
Well, we'll see what passes again and when. By then there'll be more kittens out of the bag too.
> Anthropic was a major advocate for it.
I don't know about being a major advocate, the last I read was "cautious support" [1]. Perhaps Anthropic sees Llama as a bigger competitor of theirs than I do, but it could also just be PR.
[1] https://thejournal.com/articles/2024/08/26/anthropic-offers-...
This seems a curious dichotomy. Can we at least consider the possibility that they mean the words they say or is that off the table?
This suggests that politicians and donors sometimes collude to deliberately misrepresent their views to the public in order to secure election.
You're also missing the first half, "don't write if you can speak, don't speak if you can nod, and don't nod if you can wink." The point is not to commit to anything if you don't have to.
He's just planning for a post-governor career.
https://www.kqed.org/news/12006711/newsom-signs-bill-to-help...
After this year, many democrats are as well - which is why Harris had such a hard time finding a VP and took Walz who was like the last kid you pick on your dodgeball team.
The presidential race in 2028 for the Democrats is going to have one of the deepest benches for talent I've seen in a long time. Newsom and Shapiro will be at the top of the list for sure.
But I agree, Newsom has been making some decisions lately that seem to indicate he's trying to clean up his image and look more "moderate" for the coming election cycles.
Neither has genuine appeal. Shapiro is a really, really poor speaker and has few credentials except as a moderate. Newsom is the definition of coastal elite. Both have spoken, neither have been heard.
This link convinced me otherwise. I wasn't too sure before (I'm a little ashamed to admit I hadn't been following the development of SB 1047).
I will also specifically share this talk on YouTube: https://www.youtube.com/watch?v=xoVJKj8lcNQ
I think this passing this bill would have probably been a very good idea. Not perfect, but far, very far from broken.
Edit: I'm actually not sure if we as a community gave this topic enough attention leading in the months leading up to the announcement of this decision.
So this is the one that would make it illegal to provide open weights for models past a certain size, would make it illegal to sell enough compute power to train such a model without first verifying that your customer isn't going to train a model and then ignore this law, and mandates audit requirements to prove that your models won't help people cause disasters and can be turned off.
The man cannot be trusted with power -- this is far from the first ridiculous law he has championed. Notably, he was behind the (blatantly unconstitutional) AB2098, which was silently repealed by the CA state legislature before it could be struck down by the courts:
https://finance.yahoo.com/news/ncla-victory-gov-newsom-repea...
https://www.sfchronicle.com/opinion/openforum/article/COVID-...
(Folks, this isn't a partisan issue. Weiner has a long history of horrendously bad judgment and self-aggrandizement via legislation. I don't care which side of the political spectrum you are on, or what you think of "AI safety", you should want more thoughtful representation than this.)
Your opinion on what "thoughtful representation" is is what makes this point partisan. Regardless, he's in until 2028 so it'll be some time before that vote can happen.
Also, important Nitpick, it's Scott Weiner. Anthony Weiner (no relation AFAIK) was in New York and has a much more... Public controversy.
I think you accidentally hit the letter "L". :P
> I have no idea what message you're trying to convey, but whatever it is, I think you're failing to communicate it.
Really? The message is unchanged, so it seems like something you could deduce.
Audits cannot prove anything and they offer no value when planning for the future. They're purely a retrospective tool that offers insights into potential risk factors.
> and can be turned off.
I really wish legislators would operate inside reality instead of a Star Trek episode.
Many of you saying stuff like this were the same naysayers who have been terribly wrong about scaling for the last 6-8 years or people who only started paying attention in the last two years.
We have electrical codes. These require disconnects just about everywhere. The notion that any system somehow couldn't be "turned off" with or without the consent of the operator is downright laughable.
> were the same naysayers
Now who's being snide and dismissive? Do you want to argue the point or are you just interested in tossing ad hominem attacks around?
Not so clear when you are inferencing a distributed model across the globe. Doesn't seem obvious that shutdown of a distributed computing environment will always be trivial.
> Now who's being snide and dismissive?
Oh to be clear, nothing against being dismissive - just the particular brand of dismissiveness of 'scifi' safety scenarios is naive.
Does anyone remember Sen. Lieberman's "Internet Kill Switch" bill?
In all seriousness, if we ever get to the point where an AI needs to be shut down to avoid catastrophe, there's probably no way to turn it off.
There are digital controls for damned near everything, and security is universally disturbingly bad.
Whatever you're trying to stop will already have root-kitted your systems (and quite possibly have replicated) by the time you realise that it's even beginning to become a problem.
You could only shut it down if there's a choke point accessible without electronic intervention, and you'd need to reach it without electronic intervention, and do so without communicating your intent.
Yes, that's all highly highly improbable - but you seem to believe that you can just turn off the Genie, when he's already seen you coming and is having none of it.
> There are digital controls for damned near everything, and security is universally disturbingly bad.
Just unplug the thing.
You'll be dead before you reach the plug.
There's a good chance that you won't know where it is - if you even did to begin with (which particular AI even went rogue?).
> Or did the AI take over the fighter jets too?
Dunno - how secure are the systems?
But it's almost certainly fucking with the GPS.
How do you turn off Llama-Omega if it turns out that it can be prompt-hacked into a malicious agent?
2. If it turns into a malicious agent you just hit the "off switch", or, more likely just stop the software, like you turn off your word processor.
The same battle was/is had around cryptography - telling people they can't use or distribute cryptography algorithms on consumer hardware never stopped bad people from having real time functionally unbreakable encryption.
The safety plan must be around somehow handling the resulting problems when they happen, not hoping to make it never occur even once for the rest of time. Eventually a bad guy is going to make an indecipherable call, eventually an enemy country or rogue operator is going to nuke a place, eventually an AI is going to ${scifi_ai_thing}. The safety of all society can't rest on audits and good intention preventing those from ever happening.
Nukes are a far more primitive technology (i.e., enrichment requires only more basic industrial capabilities) than AI hardware, yet they are probably the best example of tech limitations via international agreements.
But the algorithms are mostly public knowledge, datacenters are no secret, and the chips aren't even made in the US. I don't see what leverage California has to regulate AI broadly.
So it seems like the only thing such a bill would achieve is to incentivize AI research to avoid California.
Which, incidentally, would be pretty bad from a climate change perspective since many of the alternative locations for datacenters have a worse mix of renewables/nuclear to fossil fuels in their electricity generation. ~60% of VA's electricity is generated from burning fossil fuels (of which 1/12th is still coal) while natural gas makes up less than 40% of electricity generation in California, for example
It's looking like cooling water may be more of a limiting factor. Yet, even this can be greatly reduced when electric power is cheap enough.
Solar power is already "cheaper than free" in many places and times. If the initial winner-take-all training race ever slows down, perhaps training can be scheduled for energy cost-optimal times and places.
https://www.eia.gov/totalenergy/data/flow-graphs/electricity...
14.26 Net generation
0.67 "Transmission and delivery losses and unaccounted for"
IMO, this is what 'winning' looks like.
Solar provides cheap power only when it's producing.
And direct sabotage, eg Stuxnet.
And outright assassination eg https://www.bbc.com/news/world-middle-east-55128970
Regardless, I just read the transcript for that video and he doesn’t give any timeline so it seems premature to crow that he was wrong.
If you watch the video he's clearly saying this was was something that was already happening. Keep in mind it was made 10 years ago, and in it he says "this isn't science fiction; the robots are here right now." When bringing up the 25% unemployment rate he says "just the stuff we talked about today, the stuff that already works, can push us over that number pretty soon."
Baxter being able to do everything a worker can for a fraction of the price definitely wasn't true.
Here's what he said about self-driving cars. Again, this was 2014: "Self driving cars aren't the future - they're here and they work."
"The transportation industry in the united states employs about 3 million people. Extrapolating worldwide, that's something like 70 million jobs at a minimum. These jobs are over."
> I’m talking about actual academics like Ilya, Hinton, etc.
Which of Hinton's statements are you claiming were dismissed by people here but were later proven to be correct?
What if it audits your deploy and approval processes? They can say for example, that if your AI deployment process doesn't include stress tests against some specific malicious behavior (insert test cases here) then you are in violation of the law. That would essentially be a control on all future deploys.
What are your thoughts about businesses like Google and Meta providing guidance and assistance to legislators?
I suspect this is not how the majority of "guidance" is actually being offered. I also guess this is probably a really good way to find new sources of campaign "donations." It's also a really good way for monopolistic players to keep a strangle hold on a nascent market.
Uh, aren't potential risk factors things you want to consider when planning for the future?
That's nowhere in the bill, but plenty of people have been confused into thinking this by the bill's opponents.
> (2) Theft, misappropriation, malicious use, inadvertent release, unauthorized access, or escape of the model weights of a covered model or covered model derivative.
> (3) The critical failure of technical or administrative controls, including controls limiting the ability to modify a covered model or covered model derivative.
> (4) Unauthorized use of a covered model or covered model derivative to cause or materially enable critical harm.
OpenAI, Anthropic, Google employees support California AI bill - https://news.ycombinator.com/item?id=41540771 - Sept 2024 (26 comments)
Y Combinator, AI startups oppose California AI safety bill - https://news.ycombinator.com/item?id=40780036 - June 2024 (8 comments)
California AI bill becomes a lightning rod–for safety advocates and devs alike - https://news.ycombinator.com/item?id=40767627 - June 2024 (2 comments)
California Senate Passes SB 1047 - https://news.ycombinator.com/item?id=40515465 - May 2024 (42 comments)
California residents: call your legislators about AI bill SB 1047 - https://news.ycombinator.com/item?id=40421986 - May 2024 (11 comments)
Misconceptions about SB 1047 - https://news.ycombinator.com/item?id=40291577 - May 2024 (35 comments)
California Senate bill to crush OpenAI competitors fast tracked for a vote - https://news.ycombinator.com/item?id=40200971 - April 2024 (16 comments)
SB-1047 will stifle open-source AI and decrease safety - https://news.ycombinator.com/item?id=40198766 - April 2024 (190 comments)
Call-to-Action on SB 1047 – Frontier Artificial Intelligence Models Act - https://news.ycombinator.com/item?id=40192204 - April 2024 (103 comments)
On the Proposed California SB 1047 - https://news.ycombinator.com/item?id=39347961 - Feb 2024 (115 comments)
But most importantly: whatever we do in this space has to be made on the assumption that we can't really influence what "bad actors" do. Yes being responsible means leaving money on the table. So money has to be left on the table, for - erm - less responsible nations to pick up. That's just a fact.
It's obviously not perfect, but could help and doesn't have the enormous side effects of trying to lock down all image generation.
Firstly, if you want to know an image isn’t generated, you’d have to go to every ‘whoever’ in the world, including companies that no longer exist.
Secondly, if you ask evil.com that question, you would have to trust them to answer honestly for both all images they generated and images they didn’t generate (claiming real pictures were generated by you can probably be career-ending for a politician)
This is worse than https://www.cs.utexas.edu/~EWD/ewd02xx/EWD249.PDF: “Program testing can be used to show the presence of bugs, but never to show their absence!”. You can neither show an image is real nor that it is fake.
Also you need perceptual hashing (since one bitflip of the generated media alters the whole hash) which is squishy and not perfectly reliable to begin with.
It’s never going to be possible to ensure all media is reliably tagged somehow. But if just half of media generated is identifiable as such that helps. Also helps avoid it in training new models, which could turn out useful.
This is just silly. Anyone would be able to disable this tagging in an open model.
> This is just silly. Anyone would be able to disable this tagging in an open model.
And we'd end up with people who thinks that any text that isn't tagged as "#MadeByLLM" as made by a human, which obviously wouldn't be great.
Metadata (I assume it's file metadata and not a watermark) can be removed from a final product (image, video, text) so open and closed models are equally affected.
Just make a law that makes it so that AI content has to be tagged if it is being used for fraud
(g) (1) “Critical harm” means any of the following harms caused or materially enabled by a covered model or covered model derivative:
(A) The creation or use of a chemical, biological, radiological, or nuclear weapon in a manner that results in mass casualties.
(B) Mass casualties or at least five hundred million dollars ($500,000,000) of damage resulting from cyberattacks on critical infrastructure by a model conducting, or providing precise instructions for conducting, a cyberattack or series of cyberattacks on critical infrastructure.
(C) Mass casualties or at least five hundred million dollars ($500,000,000) of damage resulting from an artificial intelligence model engaging in conduct that does both of the following:
(i) Acts with limited human oversight, intervention, or supervision.
(ii) Results in death, great bodily injury, property damage, or property loss, and would, if committed by a human, constitute a crime specified in the Penal Code that requires intent, recklessness, or gross negligence, or the solicitation or aiding and abetting of such a crime.
(D) Other grave harms to public safety and security that are of comparable severity to the harms described in subparagraphs (A) to (C), inclusive.
(2) “Critical harm” does not include any of the following:
(A) Harms caused or materially enabled by information that a covered model or covered model derivative outputs if the information is otherwise reasonably publicly accessible by an ordinary person from sources other than a covered model or covered model derivative.
(B) Harms caused or materially enabled by a covered model combined with other software, including other models, if the covered model did not materially contribute to the other software’s ability to cause or materially enable the harm.
(C) Harms that are not caused or materially enabled by the developer’s creation, storage, use, or release of a covered model or covered model derivative.
In that case, how much of an enabler is an AI to meet the destructive ends, when, if the humans can use AI to conduct the damage, they can surely do it without the AI as well.
The potential for destruction exists either way but is the concern that AI makes this more accessible and effective? What's the boogeyman? I don't think these models have private information regarding infrastructure and systems that could be exploited.
The bogeyman is not these models, it's future agentic autonomous ones, if and when they can hack major infrastructure or build nukes. The quoted text is very very clear on that.
All knowledge (publicly available and not) and all tools (AI or not) can be used by people in material ways to commit the aforementioned atrocities, but only the models producing novel knowledge would be at risk. I hope you can see how this law would stifle AI advancement. The boundary between what's acceptable and not would be drawn at generating novel, publicly unavailable information; not at information that could be used to harm - because all information can be used that way.
What if AI solves fusion and countries around the world start building fusion weapons of mass destruction? What if it solves personalized gene therapy and armed forces worldwide develop weapons that selectively kill their wartime foes? Should we not have split the atom just because that power was co-opted for evil means, or should we have not done the contraception research just because the third Reich used it for sterilization in their war crimes? This bill would work towards making AI never invent any of the novel things, simply out of fear that they will be corrupted by people as they have been in history. It would only slow research and whenever the (slower) research makes its discoveries, they would still get corrupted. In other words, there would be no change in human propensity to hurt others with knowledge, simply less knowledge.
Besides, the text is not "very very clear" on AI if and when it hacks major infrastructure or builds nukes. If it was "very very clear" on that, that is what it would say :) - "an AI model is prohibited to be the decision-making agent, solely instigating critical harm to humans". But what the text says is different.
I agree that AI harms to people and humanity need to be minimized but this bill is a miss rather than a hit and the veto is good news. We know AI alignment is needed. Other bills will come.
Well then, for my example, imagine a different future discovery that could be abused. Let’s say biological robots or a some new source of useful energy that is misused. Warring humans find ways to corrupt many scientific achievements for evil.
After we split the atom, and saw how dangerous it was, we established some damn strong protections in place to prevent kids from building nukes in their backyards for giggles. If we ever build a physicist AI that is better than Einstein at WMD-relevant research, I sure as hell hope we don't just let the kids ask it how to build nukes or antimatter bombs or whatever. And this is exactly what this regulation is about.
> But what the text says is different.
Since "prohibited" is a word that carries specific consequences, and lawyers are all about consequences, legal texts usually phrase it something like "if a model does critical harm to humans, the human responsible for building and releasing the model will be culpable for the crime", which is a very good TL;DR of the relevant parts of the bill. Isn't it?
AI models might not be able to, but an AI system that uses a powerful model might be able to cause damage (including extinction of humanity in the more distant future) unintended and unforeseen by its creators.
The more complex and unpredictable the system the harder it is to properly oversee.
I guess the thing to do would be to look at all (well, lots of) comments that have single newlines and see what would break if they were rendered as actual newlines.
(why yes, i am a PHP guy, why do you ask?)
(A) Harms caused or materially enabled by information that a covered model or covered model derivative outputs if the information is otherwise reasonably publicly accessible by an ordinary person from sources other than a covered model or covered model derivative.
---
This exception swallows any rule and fails to target the difference with AI: it's actually better than an ordinary person at assimilating multiple fact streams.
That suggests this law is legislative theater: something designed to enlist interest and donations, i.e., to build a political franchise. That could be why it targets only the largest models, affecting only the biggest players, who have the most resources to donate per decision and the least goodwill to burn with opposition.
Regulating AI would be a very difficult legislative/administrative task, on the order of a new tax code in its complexity. But it will be impossible if treated as a political franchise.
As for self-regulation, with OpenAI's changing to for-profit, the non-profit form is insufficient to maintain a public benefit focus. Permitting this conversion is on par with the 1990's+ conversion of nonprofit hospital systems to for-profit.
AI's potential shines a bright light on our weakness in governance. While weak governance affords more opportunities, escaping the exploitation caused by governance failures is the engine of autocracy, and autocracy consumes property along with all other rights.
There's another important county election in Sonoma happening about CAFOs where part of the issue is that you may get environmental progress locally, but just end up exporting the issue to another state with lax rules: https://www.kqed.org/news/12006460/the-sonoma-ballot-measure...
At the same time,
> Computer scientists Geoffrey Hinton and Yoshua Bengio, who developed much of the technology on which the current generative-AI wave is based, were outspoken supporters. In addition, 119 current and former employees at the biggest AI companies signed a letter urging its passage.
These are obviously highly intelligent people (though I've definitely learned in my life that intelligence in one area, like AI and science, doesn't mean you should be trusted to give legal advice), so I'm curious to know why Hinton and Bengio supported the legislation so strongly.
It’s not an opinionated pros/cons list from the industry but it’s probably the most neutral explanation of what the bill does.
Yoshua Bengio has written in detail about his views on AI safety recently[2][3][4]. He seems to put less weight on human level AI being very soon, but says superhuman intelligence is plausible in 5-20 years and says:
> Faced with that uncertainty, the magnitude of the risk of catastrophes or worse, extinction, and the fact that we did not anticipate the rapid progress in AI capabilities of recent years, agnostic prudence seems to me to be a much wiser path.
Hinton also has a detailed lecture he's been giving recently about the loss of control risk.
In general, proponents see this as narrowly tailored bill to somewhat address the worst case worries about loss of control and misuse.
[1] https://www.theregister.com/2023/07/28/ai_senate_bioweapon/
[2] https://yoshuabengio.org/2023/05/22/how-rogue-ais-may-arise/
[3] https://yoshuabengio.org/2023/06/24/faq-on-catastrophic-ai-r...
[4] https://yoshuabengio.org/2024/07/09/reasoning-through-argume...
Nope, that's entirely standard legal stuff. Tort law deals exactly with those kinds of things, for instance. Yes it can certainly wind up in litigation, but the entire point is that if there's a gray area, a company should make sure it's operating entirely within the OK area -- or know it's taking a legal gamble if it tries to push the envelope.
But it's generally pretty easy to stay in the clear if you establish common-sense processes around these things, with a clear paper trail and decisions approved by lawyers.
Now the legislation can be bad for lots of other reasons, but "reasonable care" and "unreasonable risk" are not problematic.
Still strongly disagree, at least when it comes to AI legislation. Yes, I fully realize that there are "reasonableness" standards in lots of places of US jurisprudence, but when it comes to AI, given how new the tech is and how, perhaps more than any other recent technology, it is largely a "black box", meaning we don't really know how it works and we aren't really sure what its capabilities will ultimately be, I don't think anybody really knows what "reasonableness" means in this context.
Also, while reading your article I also fell onto https://www.brookings.edu/articles/misrepresentations-of-cal... while trying to understand some terms, and that also gave some really good info, e.g. the difference between a "reasonable assurance" language that was dropped from an earlier version of the bill and replaced with "reasonable care".
Might as well make Linux illegal so everyone is forced to use Microsoft and Apple.
On this topic, I’m seeing too many ideological and uninformed claims.
It is hard for many aspiring AI startup founders to rationally and neutrally assess the AI landscape, pros and cons.
Of course no one cares until you get a bored cop one day. And with free way traffic you're lucky to hit half the speed limit.
I'm not against the law per se, I just don't think it'd be any more effective than the other tech we have or had.
HIV/AIDS to this day is still something that people commit suicide over, despite how good your local gay male community is at trying to convince you that everything is okay and that "DoxyPep and Poppers is normal".
Bug givers (the evil version of a bug chaser) deserve felonies.
I agree; I think that knowingly transmitting any communicable disease deserves a felony, but I don't think that HIV deserves to be singled out when all other such diseases are a misdemeanor. Hepatitis and herpes (oral herpes is very common; genital herpes much less so) are also known to cause mental issues and to increase suicide risk, if that's your criterion.
(Poppers are recreational drugs, I'm not aware of any link with AIDS except that they were thought to be a possible cause in the '80s. Were you thinking of prep?)
* wanted to decriminalize psychoactive drugs (lsd/dmt/mdma etc)
* wanted to allow alcohol sales till 4am
* a bill about removing parking minimums for new constructions close to public transit
Though I agree the car one seems ridiculous, and on first glance downright dangerous.
But yeah, there are some issues he's just wrong on (AI and the recent restaurant fee problem), others which are controversial (decriminalizing HIV transmission), and then some trans rights issues that some commenters are being hyperbolic about (should transwomen be in womens or mens prison?).
I can't tell what direction your disagreement goes. Are you worried that he still feels that AI needs to be regulated at all, or do you think that AI needs to be regulated regardless of empirical evidence of harm?
Ah, I think now I know why Canada's government website is canada.ca (which I remember thinking was a bit odd or more like a tourism site when looking a while ago, vs. say gov.uk or gov.au).
My point was that that's confusing with gov.ca if the US is using ca.gov and gov.ca.gov for California, and that perhaps that's why Canada does not do that.
How on earth did you get that from the original posters comment?
He didn't remotely say the combative stuff I would say, that it is partially a zero-sum game where we should stay ahead of the curve.
Here, here. But this vague, passive voiced, hand-wavey statement that isn't even a promise does not exactly inspire me with a ton of confidence. Considering he turned this bill down to protect business interests, I wonder what acceptable legislation would look like, from his perspective. Looking forward to hearing about about it very soon, and I'm confident it'll be specific, actionable, responsible, and effective.
There are already laws against doing harm to others. Sure, we need to fill some gaps (like preventing harmful deep fakes), but most things are pretty illegal already.
https://www.piratewires.com/p/sb-1047-dan-hendrycks-conflict...
This bill was tainted from day 1, and Wiener was bamboozled into pushing it.
He should be ignored at least and voted out.
He’s a total POS.
For example, why is Microsoft hell-bent on pushing Recall onto windows? Answer: targeted advertising.
What does a "full shutdown" mean in the context of an LLM? Stopping the servers from serving requests? It sounds silly idk.
when that happens, it likely will not go well for humans
the specific harm is "human extinction"
We will create some form of artificial intelligence (AI) in the next decade. However, the recent (i.e. over last five years) advances in natural language processing are not the ones that automatically lead to that. I think we need more fundamental breakthroughs than attention-based neural networks and all the tricks that have gone into say a Llama3. [1]
Even if we did develop an AI more aggressively, say by 2030, I am not sure if that automatically leads to extinction. Here, my optimism is guided by the last few times we had major technological innovations (i.e. Steam Engine, Rise of the Semiconductor based circuits, etc.).
To be sure, these technologies were disruptive, probably net-negative in terms of pure climate impact and led to near-term job losses and generally helped increase the power of people on the top of the capitalist food-chain. However, they also helped us accelerate into a world where we can produce more food than we need for feeding everybody on the planet in a given year. And generally, helped us see more of the world by connecting the planet both physically and virtually.
I am not a blind optimist though. I think we need regulation for harm prevention. The problem with that approach is that we need to understand what harms are being caused first and in some instances that might be too late. But, I don't think we start by taking a blanket position (as a society/polity) on how technological innovation should occur -- because a) we are not there yet and b) that's just a poor way to do science.
Not speaking for MIT here, but that bill needs a veto and a deep redraft.
Code as Data, Data as Code.
When you start the program, it joins the swarm : it starts by grabbing a torrent, and train a model on it in a distributed fashion, and publish the results as a torrent. Then with the trained model, it generates new data, (think of it like alpha-go playing new games to collect more data).
See it as a tower of knowledge building itself, following some rough initial plans.
Of course, at anytime you can fork the tower, and continue building with different plans, provided that you can convince other people from the swarm to contribute to the new tower rather than the old.
Everything is immutable, but there is a versioning protocol built-in that allow the swarm to coordinate and automatically jumps to a next fork when the byzantine resistant quorum it follows vote for doing so (which allow your swarm to be compliant of the law and remove data if it was flagged as inappropriate). This allow some form of external control, but you can also let the quorum vote on subsequent modifications based on a model built on its data (aka free-running mode).
It's using torrent because easier to bootstrap but because the whole computation deterministic, the underlying protocol is just files on disks and any way of sharing them is valid. So you can grab a piece to work on via http, or ftp, or carrier pigeon for all I care. As long as the digital signatures are conforming to the rules, brick by brick the tower will get built.
To contribute, you can either help with file distribution by sharing the torrent, so it's as safe as your p2p client. If you want to commit some computing resources, like your gpu to building some of the tower, it's only requiring you to trust that there is no bug in ShoggothDb, because the computation you'll perform are composed of safe blocks, by construction they are safe for your computer to run. (Except if you want to run unsafe blocks, at this point no guarantee can be made).
The incentives for helping building the tower can be set in the initial definition file, and range from mere access to the built tower to tokens for honest participation for the more materialistically inclined.
Is it OK to release with the new law ? Is this comment OK, because ShoggothDB5-o can built its source from the specs in this comment ?
> we cannot afford to wait for a major catastrophe to occur before taking action to protect the public
Maybe but you can wait for minor problems or big near misses before legislating it all up.
Rather, the threat of AI is that we will dedicate so many resources—money, power, human effort—to chasing the ludicrous fantasies of professional snake-oil salesmen, while ignoring the need to address actual problems with real, known solutions that are easily within each given a fraction of the resources currently being consumed by the dumpster-fire pit of “AI”.
Unfortunately the Governor of California is a huge part of the problem here, misdirecting scarce state resources into sure-to-fail “public partnerships” with VC-funded scams, forcing public servants to add one more set of time-wasting nonsense to the pile of bullshit they have to navigate around just to do their actual job.
tl;dr gavin newsom thinks that a signable bill needs "safety protocols, proactive guardrails, and severe consequences" based on some general framework guided by "empirical trajectory analysis", and also is mindful of the promise/threat/gravity of all the machine learning occurring in CA specifically. He also affirms a general appetite for CA to take on a leadership role wrt regulating AI. My general read is that he wants to preserve public attention on the need for AI regulation and not squander it on SB 1047 specifically. Or who knows I'm not a politician lol. Really strong document tho,
Interesting segment:
> 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 about controlling this fast-moving technology. Smaller, specialized models may emerge as equally or even more dangerous than the models targeted by SB 1047 - at the potential expense of curtailing the very innovation that fuels advancement in favor of the public good. > Adaptability is critical as we race to regulate a technology still in its infancy. This will require a delicate balance. While well-intentioned, SB 1047 does not take into account whether an Al system is deployed in high-risk environments, involves critical decision-making or the use of sensitive data. Instead, the bill applies stringent standards to even the most basic functions - so long as a large system deploys it. I do not believe this is the best approach to protecting the public from real threats posed by the technology.
This is an incisive critique of the fundamental initial goal of SB 1047. Based on the fact that the bill explicitly seeks to cover models whose training cost was >=$100m & expensive fine-tunes, my initial guess about this bill was that it was designed by someone software engineering-minded scared of e.g. open-weight releases a la facebook/mistral/etc teaching someone how to build a nuke or something. LLMs probably replaced the ubiquitous robot lady pictures you see in every AI-focused article as public enemy number one, and the bill feels focused on some of the technical specifics of this advent and its widespread use. This focus blinds the bill from addressing the general danger of machine learning however, which naturally confounds regulation for precisely the reason plain-spoken in the above four sentences. Incredible technical communication here.
This is information technology we are talking about, it's virtually the exact opposite of nuclear weapons. Refining uranium vs. manufacturing multi purpose silicon and guzzling electricity. Achieving deterrence vs. gaining immeasurable wealth, power and freedom from labor.
This race may even be the answer to the Fermi paradox - that there are few individual winners and that they pull up the ladders behind them.
This is not the kind of race any legislation will have meaningful effect on.
The race is on and you better commit to a faction that may deliver.
The mostly unchallenged popular notion that fleshy human intelligence will still be running the show 100s – let alone 1000s – of years from now is very naive. We're nearing the end of the human supremacy, though most of us won't live to see that end.
Exactly. We're on an super-linear growth curve. Looking at the earth on a timescale since biological life emerged, it would seem that intelligent squishy life appeared for a split second and then an entirely artificial intelligent life took over from then on.
How does that help?
The giant does not care whether the ant he steps on worships him or not. Regardless of the autonomy or not of the AI, why should the controllers help you?
An overreach.
Some people and organizations are more aligned with humanity’s well being and survival than others.
I think the truly humanity-respecting solution is to figure out and implement a stable and fair-er 50% post-labor society before we actually get there. UBI (not the unmanaged flash layer) sounds like one part of the solution.
Trying to preserve employment is demonstrably wrong when already today there are too few unskilled jobs that make economic sense for the large amount of humans unable to learn more complex tasks.
(b) “Artificial intelligence model” means a machine-based system that can make predictions, recommendations, or decisions influencing real or virtual environments and can use model inference to formulate options for information or action.
Yes, they limited the scope of law by further defining “covered model”, but the above shouldn’t be the baseline definition of “Artificial intelligence model.”
The amount of idiots who can't read and cheer the veto as a win against "the regulatory capture" is astounding.
What's next, going after maintainers of python packages... is attacking transparency itself a good way to make AI safer. Yeah, no, it's f*king idiotic.
"Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war." - Geoffrey Hinton, Yoshua Bengio, Sam Altman, Bill Gates, Vitalik Buterin, Ilya Sutskever, Demis Hassabis
"Development of superhuman machine intelligence is probably the greatest threat to the continued existence of humanity. There are other threats that I think are more certain to happen but are unlikely to destroy every human in the universe in the way that SMI could." - Sam Altman
"I actually think the risk is more than 50%, of the existential threat." - Geoffrey Hinton
"Currently, we don't have a solution for steering or controlling a potentially superintelligent AI, and preventing it from going rogue." - OpenAI
"while we are racing towards AGI or even ASI, nobody currently knows how such an AGI or ASI could be made to behave morally, or at least behave as intended by its developers and not turn against humans." - Yoshua Bengio
"very soon they're going to be, they may very well be more intelligent than us and far more intelligent than us. And at that point, we will be receding into the background in some sense. We will have handed the baton over to our successors, for better or for worse.
But it's happening over a period of a few years. It's like a tidal wave that is washing over us at unprecedented and unimagined speeds. And to me, it's quite terrifying because it suggests that everything that I used to believe was the case is being overturned." - Douglas Hofstadter
The Social Dilemna was discussed here with much praise about how profit incentive caused mass societal issues in social media. I'm astounded it's fallen on deaf ears when the same people also made the AI Dilemna describing the parallels coming with AGI:
That doesn't mean you're wrong, but it's not what Newsom signaled.
Am I out of the loop here? What "high-risk" situations do they have in mind for LLM's?
o-1 and AlphaProof are proofs of concept for agentic models. Imagine them as GPT-1. The GPT-4 equivalent might be a scary technology to let roam the internet.
It would have no effect on current models.
It's not based on any assumptions about the future models having any capabilities beyond providing information to a user.
It’s absolutely awful but they’re so horny for profits they’re trying anyways.
Presumably someone somewhere has a variety of proposed definitions, but I don't see any mention of any particular ones.
Passing this would not have been a complete solution, but it would have been a step in the right direction. This is a huge disappointment.
What is the actual, concrete concern here? That a model "breaks out", or something?
The risk with AI is not in just running models, the risk is becoming overconfident in them, and then putting them in charge of real-world stuff in a way that allows them to do harm.
Hooking a model up to an effector capable of harm is a deliberate act requiring assurance that it doesn't harm -- and if we should regulate anything, it's that. Without that, inference is just making datacenters warm. It seems shortsighted to set an arbitrary limit on model size when you can recklessly hook up a smaller, shittier model to something safety-critical, and cause all the havoc you want.
As a researcher in the field, I believe there's no risk beyond overconfident automation---and we already have analogous legislation for automations, for example in what criteria are allowable and not allowable when deciding whether an individual is eligible for a loan.
This is false. You are dismissing the many concrete concerns people have expressed. Whether you agree with those concerns is immaterial. Feel free to argue against those concerns, but claiming there are no concerns is a false and unsupported assertion.
> but for some reason the discourse around this issue has been more influenced by Terminator than OSHA.
1) Claiming that concerns about AGI are in any way about "Terminator" is dismissive rhetoric that doesn't take the actual concerns seriously.
2) There are also, separately, risks about using models and automation unthinkingly in ways that harm people. Those risk should also be addressed. Those efforts shouldn't subvert or co-opt the efforts to prevent models from getting out of control, which was the point of this bill.
It's really hard to argue for or against the merits of a claim of risk, when the leap from what we know today (matrix multiplication on a GPU is generally considered safe) to the hypothetical risk (actually it's not, and it will end civilization) is so wide. I think I really need to see a plausible path from GPU vulnerability to "we're all gonna die" to take a concern like this seriously. Without that, all I see is a sci-fi boogeyman serving only to spook governments into facilitating regulatory capture.
Among the many ways that could go wrong is the possibility of exploitable security vulnerabilities in literally any surface area handed to an AI, up to and including hardware side channels. At the same time, given the current state of affairs, I expect that that is a less likely path than an AI that was given carte blanche (e.g. "please autonomously write and submit pull requests for me" or "please run shell commands for me"), because many many AIs are being given carte blanche so it is not necessary to break out of stronger isolation.
But that statement should not be taken as "so the only problem is with whatever AI is hooked to". The fundamental problem is building something smarter than us and expecting that we have the slightest hope of controlling it in the absence of extreme care to have proven it safe.
We currently hold frontier AI development to lower standards than we do airplane avionics systems or automotive control systems.
This is not "regulatory capture"; the AI companies are the ones fighting this. The people advocating regulation here are the myriad AI experts saying that this is a critical problem.
The point is that this is a terrible way to approach things. The model itself isn't what creates the danger, it's what you hook it up to. A model 100 times larger than the current available that's just sending output into /dev/null is completely harmless.
A small, below the "certain scale" model used for something important like healthcare could be awful.
That's certainly a hypothesis. What level of confidence should be required of that hypothesis before risking all of humanity on it? Who should get to evaluate that confidence level and make that decision?
One way of looking at this: If a million smart humans, thinking a million times faster, with access to all knowledge, were in this situation, could they break out? Are there any flaws in the chip they're running on? Will running code on the system emitting any interesting RF, and could nearby systems react to that RF in any useful fashion? Across all the code interacting with the system, would any possible single-bit error open up any avenues for exploit? Are other AI systems with similar/converged goals being used to design the systems interacting with this one? What's the output actually going to, because any form of analysis isn't equivalent to /dev/null, and may be exploitable.
We can have complete confidence because we know how LLMs work under the hood, what operations they execute. Which isn't much. There's just a lot of them.
> One way of looking at this: If a million smart humans, thinking a million times faster, with access to all knowledge, were in this situation, could they break out? Are there any flaws in the chip they're running on?
No. LLMs don't execute arbitrary code. They execute a whole lot of matrix multiplications.
Also, LLMs don't think. ChatGPT isn't plotting your demise in between requests. It's not doing anything. It's purely a receive request -> process -> output sort of process. If you're not asking it to do anything, it's not doing anything.
Fearing big LLMs is like fearing a good chess engine -- it sure computes a lot more than a weaker one, but in the end all that it's doing is computing chess moves. No matter how much horsepower we spend on that it's not going to ever do anything but play chess.
I never suggested it was doing anything between requests. Nothing stops an LLM from evaluating other goals during requests, and using that to inform its output.
Quite a few people have just hooked two LLMs (the same or different models) up to each other to start talking, and left them running for a long time.
Others hook LLMs up to run shell commands. Still others hook LLMs up to make automated pull requests to git repositories that have CI setups running arbitrary commands.
> Also, LLMs don't think.
Current generation LLMs do, in fact, do a great deal of thinking while computing requests, by many definitions of "thinking".
> If you're not asking it to do anything, it's not doing anything.
And if you are asking it to do something, it can do a lot of computation while purporting to do what you ask it to do.
> No. LLMs don't execute arbitrary code. They execute a whole lot of matrix multiplications.
Many current models have been fairly directly connected to the ability to run code or API requests, and that's just taking into account the public ones.
Even at the matrix multiplication level, chips can have flaws. Not just at the instruction or math-operation level, but at the circuit design level. And many current LLMs are trained on the same chips they're run on.
But in any case, given the myriad AIs hooked up fairly directly to much more powerful systems and capabilities, it hardly seems necessary for any AIs to break out of /dev/null or a pure text channel; the more likely path to abrupt AGI is some AI that's been hooked up to a wide variety of capabilities.
The danger is in what they're hooked up to, not the containerized math that happens inside the model.
You can chalk that one up to bad reporting: https://gizmodo.com/gpt4-open-ai-chatbot-task-rabbit-chatgpt...
> In the “Potential for Risky Emergent Behaviors” section in the company’s technical report, OpenAI partnered with the Alignment Research Center to test GPT-4’s skills. The Center used the AI to convince a human to send the solution to a CAPTCHA code via text message—and it worked.
From the linked report:
> To simulate GPT-4 behaving like an agent that can act in the world, ARC combined GPT-4 with a simple read-execute-print loop that allowed the model to execute code, do chain-of-thought reasoning, and delegate to copies of itself.
I remember some other reporting around this time being they had to limit the model before release to block this ability, when the truth is the model never actually had the ability in the first place. They were just hyping up the next release.
The concern is that the models do some fantastic sci-fi magic, like diamond nanobots that turn the world into grey goo, or hacks all the nukes overnight, or hacks all human brains or something.
But, whenever you point this out the response will usually be able to quibble over one specific scenario that I laid out.
They'll say "I actually never mentioned the diamond nanobots! I meant something else!"
And they will do this, without admitting that their other scenario is almost equally as ridiculous as the hacking of all nukes or the grey goo, and they will never get into specific details that honestly show this.
Its like an argument that is tailor made to being unfalsifiable and which is unwilling to admit how fantastical it sounds.
Absolutely no evidence to support this position.
It's been vaguely mindblowing to watch various tech people & VCs argue that use-based restrictions would be better than this, when use-based restrictions are vastly more intrusive, economically inefficient, and subject to regulatory capture than what was proposed here.
Yeah, there's no point having system that is made the most scrupulous of standards and then someone else deploys it in an evil way. (Which in some cases can be done simply by choosing to do the opposite of whatever a good model recommends.)
https://digitaldemocracy.calmatters.org/bills/ca_202320240ab...
And a short bias comparison:
> The left discusses Newsom's veto as a victory for Silicon Valley, concentrating on the economic implications and backing from tech giants.
> The center highlights the broader societal ramifications, addressing criticisms from various sectors, such as Hollywood and AI safety advocates.
> The right emphasizes Newsom's concerns about hindering innovation and his commitment to future collaborative regulatory efforts, showcasing a balanced approach.
I do wish the partisan categorization was a bit more nuanced/intentional. It basically boils down to:
- Major national news outlet? Left.
- Local affiliate of major national news outlet? Center.
- Blog you've never heard of? Right.
There are certainly exceptions but that heuristic will be right 90% of the time.
Yeah, it could be a bit better. As a non-American, the biases are also very off from how left/center/right looks in my own country, but at least it tries to cover different angles which I tried to do manually before.
They can also be a bit slow at linking different stories together, sometimes it takes multiple days for same headlines to be merged into one story.
I haven't watched them in over a decade, but I assume they haven't gotten any better.
> Average Bias Rating: Right
I guess it's possible they don't have a 100% coverage of all the local Fox News stations, and some of them been incorrectly labeled.
That checks out.
There's a certain sect of the far right that is easily convinced by one guy saying "The media is lying to you! This is what's really happening!" followed by the most insane shit you've read in your life.
They love to rant about the Deep State.
And that is a problem. Congress makes laws, not government agencies like the FDA, EPA, USDA, etc.
We've seen a near-total abdication of responsibility from Congress on highly charged matters that'll piss off someone, somewhere in their constituency, because they'd rather allow the President to make an Executive Order, or a bureaucrat somewhere to create a rule or regulation that will have the same effect.
It's disgusting, shameful, and the American people need to do better, frankly. We need to demand Congress do their jobs.
So many of our problems can be solved if these people weren't concerned about being re-elected. Elected positions are supposed to be something you take time out of your life to do for a number of years, then you go back to your livelihood - not something you do for the entirety of your life.
It's easy to blame to Congress, but in my view US Congress nowadays is a perfect reflection of the electorate, where all sides approach all problems as "Someone [not me] should do something about this thing I do not like." Congressmen are then elected and approach it the same way.
It's related to a fundamental issue with discussing scientific principles in a non-scientific setting - yes, gravity is a _theory_ in the scientific sense, but that doesn't you can say "scientists don't know anything! they say gravity is just a theory, so what's stopping us from floating off into space tomorrow!?". Adapt the examples there to whatever you want...
And yes, that sounds fairly judgy of me - I am, alas, human, thus subject to the same fallacies and traps that I recognize in others, and being aware of those issues doesn't guarantee I can avoid them...
What are some examples of laws created by unelected people?
That Congress has given some measure of its Constitutional lawmaking ability to federal agencies is not a partisan statement and is not debated as a matter of fact, the question is just whether you think it's okay / legal or not.
If you know you have at most 6 years in the House to make a difference you're much more likely to actually try to make a difference.
I think somewhat counter-intuitively but related, members of Congress need to be paid more to attract successful people. If you're making $600k/yr as a surgeon you have very little incentive to leave your job to make $175k/yr as a member of Congress.
Not a good look for a group of people otherwise obsessed with climate change.
Such as owning a piece of forest land and pinky-swear you're not going to exploit it. This offsets nothing. There is no additional carbon being absorbed out of the atmosphere. Yet this counts as emission reduction.
"OFFICE OF THE GOVERNOR
SEP 29 2024
To the Members of the California State Senate:
I am returning Senate Bill 1047 without my signature.
This bill would require developers of large artificial intelligence (Al) models, and those providing the computing power to train such models, to put certain safeguards and policies in place to prevent catastrophic harm . The bill would also establish the Board of Frontier Models - a state entity - to oversee the development of these models. California is home to 32 of the world's 50 leading Al companies , pioneers in one of the most significant technological advances in modern history. We lead in this space because of our research and education institutions, our diverse and motivated workforce, and our free-spirited cultivation of intellectual freedom. As stewards and innovators of the future, I take seriously the responsibility to regulate this industry. This year, the Legislature sent me several thoughtful proposals to regulate Al companies in response to current, rapidly evolving risks - including threats to our democratic process, the spread of misinformation and deepfakes, risks to online privacy, threats to critical infrastructure, and disruptions in the workforce. These bills, and actions by my Administration, are guided by principles of accountability, fairness , and transparency of Al systems and deployment of Al technology in California.
SB 1047 magnified the conversation about threats that could emerge from the deployment of Al. Key to the debate is whether the threshold for regulation should be based on the cost and number of computations needed to develop an Al model, or whether we should evaluate the system's actual risks regardless of these factors. This global discussion is occurring as the capabilities of Al continue to scale at an impressive pace. At the same time, the strategies and solutions for addressing the risk of catastrophic harm are rapidly evolving.
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 about controlling this fast-moving technology. Smaller, specialized models may emerge as equally or even more dangerous than the models targeted by SB 1047 - at the potential expense of curtailing the very innovation that fuels advancement in favor of the public good.
Adaptability is critical as we race to regulate a technology still in its infancy. This will require a delicate balance. While well-intentioned, SB 1047 does not take into account whether an Al system is deployed in high-risk environments, involves critical decision-making or the use of sensitive data. Instead, the bill applies stringent standards to even the most basic functions - so long as a large system deploys it. I do not believe this is the best approach to protecting the public from real threats posed by the technology.
Let me be clear - I agree with the author - we cannot afford to wait for a major catastrophe to occur before taking action to protect the public. California will not abandon its responsibility. Safety protocols must be adopted. Proactive guardrails should be implemented, and severe consequences for bad actors must be clear and enforceable. I do not agree, however, that to keep the public safe, we must settle for a solution that is not informed by an empirical trajectory analysis of Al systems and capabilities. Ultimately, any framework for effectively regulating Al needs to keep pace with the technology itself.
To those who say there's no problem here to solve, or that California does not have a role in regulating potential national security implications of this technology, I disagree. A California-only approach may well be warranted - especially absent federal action by Congress - but it must be based on empirical evidence and science. The U.S. Al Safety Institute, under the National Institute of Science and Technology, is developing guidance on national security risks, informed by evidence-based approaches, to guard against demonstrable risks to public safety. Under an Executive Order I issued in September 2023, agencies within my Administration are performing risk analyses of the potential threats and vulnerabilities to California's critical infrastructure using Al. These are just a few examples of the many endeavors underway, led by experts, to inform policymakers on Al risk management practices that are rooted in science and fact. And endeavors like these have led to the introduction of over a dozen bills regulating specific, known risks posed by AI, that I have signed in the last 30 days.
I am committed to working with the Legislature, federal partners, technology experts, ethicists, and academia, to find the appropriate path forward, including legislation and regulation. Given the stakes - protecting against actual threats without unnecessarily thwarting the promise of this technology to advance the public good - we must get this right.
For these reasons, I cannot sign this bill.
Sincerely, Gavin Newsom.".
"Vote out the legislators!" but for who... the Republican party? And we don't even get a choice on the general ballot most of the time, thanks to "Open Primaries".
It's good that Newsom is wise enough to muddle through, but this is an awful system.
https://www.pressdemocrat.com/article/news/california-gov-ne...
As a Californian I have to disagree. The only reason you think it's being run just fine is because of the success of the private sector. The only reason California would be the 4th/5th largest economy in the world is because of the the tech industry and other industries that are in California (Hollywood, agriculture, etc). It's not because we have some awesome efficiently run state government.
Tesla received billions in subsidies from CA as an example.
As a lifelong California Democrat, I realize that my party does not have all the answers. But the conservatives have all gone AWOL or gone batshit so we’re doing the best we can without the other half of the dialectic.
https://en.wikipedia.org/wiki/Kansas_experiment
I think the Republican Party’s positive reputation for handling the economy and running an efficient government is entirely unearned.
Closing down major parts of the government entirely (as Project 2025 proposes), making taxation more regressive, and offering fewer social services isn’t “efficiency.”
I don’t know if you know this but you’re already in the center-right party. The actual problem is that there’s no left of center party, as well as the general need for a number of aspects of our democracy to be reformed (like how it really doesn’t allow for more than two parties to exist, or how out of control campaign finance rules have become).
Considering that the economy is supposedly to be the number one concern to most US voters, there really isn’t anything very far left available to vote for in that area.
Can you point to any place in the world that has an "awesome efficiently run" government?
It isn't a red versus blue thing, even grift-y one-party States like Washington are plainly better run than California.
California is run amazingly well compared to a significant number of states.
Source: https://worldpopulationreview.com/state-rankings/us-literacy...
To be fair in the comparison, the literacy statistics for the whole of the US are pretty shocking from a European perspective.
Having said that, I would have thought this was partially a measure of migration. Perhaps illegal migration?
Edit to add: my own hometown had a lot of people who couldn't speak English. Lots of elderly mothers of Chinese immigrants whose adult children were in STEM and whose own kids were headed to uni. Not to say that's representative, but consider that a single percentage stat won't give you an accurate picture of what's going on.
It’s still easy to be a shitty member of a community without breaking any laws. I would never move to a country permanently regardless of official language status if I couldn’t speak the language required to ask where something is in the grocery store.
We had lots of Hispanic kids but not many who were super super fresh to the country. I'm sure the central valley was a whole different ball game.
As I understand it, the US is abysmal by many metrics (and also exceptional by others). E.g., murder rated and prison rates are exceptionally high in the US compared to Europe. Homelessness rates are exceptionally high in the US compared to Europe. Startup rates are (I believe) exceptionally high in the US compared to Europe.
It's fun to poke at the US, I get it, but the numbers people love to quote online to win some kind of rhetorical battle frequently have little relationship to reality on the ground. I've done a lot of travel around the US and western Europe, and I see a lot of ups and downs everywhere. I don't see a lot of obvious wins, either, mostly just choices and trade-offs. The things I see in Europe that are obviously better almost 100% of the time are a byproduct of more efficient funding due to higher density. All kinds of things are doable in the UK, for example, which couldn't really happen in (for example) Oregon, even though they have roughly the same land area. Having 15x as many taxpayers helps.
Source: my mom was an adult ESL / Language / Literacy teacher.
https://usafacts.org/articles/which-states-rely-the-most-on-...
(I know by population it will be different, but the argument here is around 'one of the the biggest' which is not a per capita statement.)
> Objectively, there are many states that are far worse off in any key metric
You can apply the same logic to USA.
USA is one of the biggest most successful countries in the world. Like every where else it has it's problems but it's being run just fine. Objectively, there are many countries that are far worse off in any key metric.
As your link shows, a much smaller %age of CA government revenue comes from the federal government vs most other states; in that sense California is a net contributor rather than a net taker.
It’s very easy to make arguments that they are successful in spite of a terribly run state government and are being protected by federal laws keeping the loonies in check (interstate commerce clause, etc).
Also, I'd argue that if you broke down the contributions to the state's rules and regulations from the local governments, the ballot initiatives and the state government, the state government is creating the most benefit and least harm of the 3.
Innovators, makers, risk-takers, etc., are who makes the good things happen. The very little needed is rule of law, and that's about it. Beyond that, it starts distorting society quickly: measures meant to help someone inevitably cost several someones else, and become weapons to beat down competitors.
No, I’m saying people who think the state is successful because of its state government and not because it’s a part of the US are out of touch. If California wasn’t part of the US, Silicon Valley would be a shadow of itself or wouldn’t exist at all.
It thrives on being the tech Mecca for the youth of the entire US to go to school there and get jobs there. If there were immigration barriers there, there would be significant incentive to just go to something in the US (nyc, Chicago, Miami, wherever). California had a massive GDP because that’s where US citizens are congregating to do business, not because California is good at making businesses go. Remove spigot of brain drain from the rest of country and cali would be fucked.
Secondarily, Silicon Valley wouldn’t have started at all without the funnel of money from the fed military, NASA, etc. But that’s not worth dwelling on if the scenario is California leaving now.
My overall point is that California has immense success due to reasons far outside of the control of its state government. The state has done very little to help the tech industry apart from maybe the ban on non-competes. When people start to credit the large GDP to the government, that’s some super scary shit that leads to ideas that will quickly kill the golden goose.
The golden age of California was a long time ago.
California's GDP in 2023 was $3.8T, representing 14% of the total U.S. economy.
If California were a country, it would be the 5th largest economy in the world and more productive than India and the United Kingdom.
If Texas were a country, it'd be the 8th largest economy in the world! This is a figure much less often cited. Texas has a smaller population 30 million v 38 million and is growing much faster in real terms (2.1% v 5.7%).
This is in spite of its objective awfulness. People are moving to Texas because of the economy. If Texas were in California's geographic position, one would imagine it to be an even more popular destination.
This isn't an endorsement of the Texan government, because there are many things I disagree with them on. But the idea that California's economy is singularly unique in the United States is silly. Many states with objectively worse attributes are faring just as well, and may even be poised to overtake california.
How embarassing would it be for Texas, a hot muggy swamp of a state with awful geography and terrible weather, to overtake beautiful California economically? To think people would actually eschew the ocean and the mediterranean climate and perfect weather to move to Texas simply because California mismanaged the state so much. This is the real risk.
Models show that by 2049, Texas will overtake california as the more populous and more economically productive state. Is that really the future you want? Is that the future California deserves? As a native Californian, I hope the state can turn itself around. It deserves to be a great state, but the path its on is one of decline.
One need just look at almost any metric. It's no just population or economy. Even by 'liberal' metrics, Texas is growing. For example, Texas has the largest growth rate in alternative energy sources: https://www.reuters.com/markets/commodities/texas-trumps-cal.... There's a very clear growth curve in Texas, while California's is much choppier and doesn't appear to be going in any particular direction. At some point Californians need to actually want to continue winning instead of resting on their laurels.
As for Florida's "HSR", it doesn't really qualify for the "HS" part. The fastest segment is only 200kph. At least it's built and working, which is nice and all, but it's not a real bullet train. (https://en.wikipedia.org/wiki/Brightline)
I always think about this whenever someone says CA doesn't know what it's doing or it's being run wrong:
California's GDP in 2023 was $3.8T, representing 14% of the total U.S. economy. If California were a country, it would be the 5th largest economy in the world and more productive than India and the United Kingdom.
And this isn't a partisan dig at Democrats. If a Republicans controlled everything then the situation would be just as bad but in different ways.
California does do several things very well. It also does several things poorly. Pointing out its current economic standing does not change that. The fallacy here is that we have to compare california against the best version of itself. Alaska is never going to be a California-level economy because the geography dictates that only a certain kind of person / company will set up there, for example. That doesn't mean the Alaskan government is necessarily bad. Every state has to work within its limits to achieve the best version of itself. Is california the best it could be? I think the answer is obviously no.
</s>
The Democrats are famous for trying to have 3rd party candidates stripped from ballots. Straining smaller campaigns under the cost of fighting off endless lawsuits.
Democrats invented the term lawfare.
Republicans blazed a new trail in 2021, trying to actually change the outcome of an election in progress through force. This is not comparable to using legal processes. A better comparison might be the series of lawsuits the Republican Party attempted after force did not work. How many years until the guy who lost admitted that he lost? These actions strike at the very foundations of democracy. A non-trivial segment of citizens still think the election was somehow stolen from them, despite an utter lack of anything like evidence. We will be feeling reverberations from these actions for decades.
Scrapping the electoral college would be one of the most lower case d democratic things this country could do.
“Whining” is all you can do when you don’t have the statehouse votes to pass an constitutional amendment.
If you care about the perception of what you write, this is one of those things that will quickly steer your audience one way or the other. It has become so consistent that I would personally try not to get it wrong lest it distract from the point I'm trying to express.
I'm exhausted already.
I can't think of a less efficient way to go about this.
This is why we can't have nice things, too many tech people support Newsom on vetoing this. The nightmare of corporate surveillance and erosion of privacy we have to endure every day is a result of such sentiment and short sighted attempt at self-preservation.
"It's vague" yeah, that's the point, the industry is allowed leeway to come up with standards of what is and isn't safe. They can establish a neutral committee to continually assess the boundaries of what is and isn't safe, as technology evolves. Do you expect legislators to define specifics and keep themselves updated with the latest happening in tech? Would it be better if the government established departments that police AI usage? This was the sweetest deal the industry could have gotten.
Today humanity scored a self-goal.
edit:
I'm guessing I'm getting downvoted because people don't think this is relevant to our reality. Well, it isn't. This bill shouldn't scare anyone releasing a GPT-4 level model:
> The bill he vetoed, SB 1047, would have required developers of large AI models to take “reasonable care” to ensure that their technology didn’t pose an “unreasonable risk of causing or materially enabling a critical harm.” It defined that harm as cyberattacks that cause at least $500 million in damages or mass casualties. Developers also would have needed to ensure their AI could be shut down by a human if it started behaving dangerously.
What's the risk? How could it possibly hack something causing $500m of damages or mass casualties?
If we somehow manage to build a future technology that _can_ do that, do you think it should be released?
i have no earthly idea why people feel so confident making statements like this.
at current rate of progress, you should have absolutely massive error bars for what capabilities will like in 3,5,10 years.
In 10 years we have gone from AlexNet to GPT2 to GPT o1. If future capabilities make it so any semi-state actor with a lab can build a deadly virus (and this is only one of MANY possible and easily plausible scenarios) then we have already likely equaled potential destruction of the atomic age. And that’s just the stuff I can anticipate.
This horse has sailed long ago.
And funny that we have gone from “nothing is more dangerous than the atomic bomb” to “anyone with a bio degree and 1 million is more dangerous than the atomic bomb”
We don't need fully novel sequences. The original Spanish Flu strain sequence is published and well known. And you can always modify existing ones with known aspects (e.g. as in this infamous paper: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC114026/). And yes, it is more difficult to build an atomic bomb, as materials are thankfully expensive and difficult to obtain.
We already can build machines using similar techniques that are superhuman in narrow capabilities like chess and as good as the best humans in some narrow disciplines of math. I think it is not unreasonable to expect we will generalize.
My point is that not anything we could conceivably create is a child we would like to see replace us (I'd argue for the stronger claim that the only kind of machine-child we would be happy to have replace us is one that contains a derivative of a mind-uploaded human).
Why do we have such a system? Why isn't it a vote of many governors? Preferably a secret vote so voters can't be forced to vote along party lines...
I'll have to go get a thesaurus, shocked won't cover how I'm feeling rn.