OpenAI, Anthropic, Google employees support California AI bill | Hacker News Reader

OpenAI, Anthropic, Google employees support California AI bill

axios.com

20 pointselsewhen26 comments

In the definitions of which models it applies to:

> An artificial intelligence model trained using a quantity of computing power greater than 10^26 integer or floating-point operations, the cost of which exceeds one hundred million dollars ($100,000,000) when calculated using the average market prices of cloud compute at the start of training as reasonably assessed by the developer

So unless you casually think you're going to throw 100 million dollars at a model training run, you're safe to innovate.

The bill itself seems reasonable to me and I'm in the camp of superalignment not really mattering because you don't know who the AI is going to align with, and there's many people with bad intent. The outcome of a model deciding to be evil or an evil person using a super-aligned model is pretty much the same except for 1 extra prompt.

How exactly are they going to measure the compute number?

> when calculated using the average market prices of cloud compute at the start of training as reasonably assessed by the developer

1. The point is once a bill like this goes through, it is easy to lower the numbers.

2. And in the future, these operations will be much cheaper.

3. New or better architectures can be more powerful with lesser compute.

If the operations are getting cheaper over time, doesn't that leave more room for AI to get better as time goes on?

Yes, making the bill, as it is now, useless over time.

And in the present, it adds unnecessary constraints for California companies while companies outside CA/US don't face such issues.

The same amount of compute will cost ~$1500 after 16 doublings. In 24 years or less.

In a previous version of the bill it was either that many flops or that much money. Now it's both.

Was it very shocking when you woke up today and found yourself transported from the year 2000, when Moore's Law still held?

Your comment is a bit snarky and I understand what you're saying but my understanding was that everyone thought Moore's Law was over when the computers stopped getting 2x faster every 18 months but that's not what Moore said. He said the transistor count would double amd it just so happened that that meant computers got a lot faster too because they benefited from all the xtra transistors. But GPUs do still get faster with more transistors because of the embarrassingly parallel nature of many matrix operations. So doubling the transistor count could still effectively significantly improve the speed and reduce time of AI training which is dominated by matrix math. Am I wrong? Did the transistor count doubling also slow down?

There's always someone out there claiming Moore's law to be dead. It never is. Chiplets and chip stacking in particular have enabled big increases in transistor count without associated downsides like decreased yield. So long as there is innovation and competition in the semiconductor industry, density / $, which you correctly point out was Moore's original metric, will continue to increase.

The potential for increasing performance and decreasing cost of LLM inference is even higher than Moore's law suggests, as the algorithms improve and are pushed closer to ideal hardware to run them.

How many doublings are needed until we harness the entirety of observable universe into building a huge computer?

Doesn't it cement the position of these companies through regulatory capture?

It doesn't even need to be regulatory capture to cement their positions and prevent competition. It just needs to be enough regulatory burden (even if justified) to prevent competitors (or other human persons) from doing what they did.

I'd be shocked if the megacorps were not actively lobbying to seal off the market like this.

Slightly off-topic: Is it legal to call your company whatever you want in the US? For example, could I create a company called "Free food for everyone" but then charge for the food I'm offering? Or a company called "The only non-poisoned cupcakes in NYC"?

Seemingly, OpenAI does everything they can do prevent competitors or helping anyone else but their investors/shareholders, but at one point someone has to have enough influence/smarts to get them to change the name, right?

> "The only non-poisoned cupcakes in NYC"?

The issue would more likely be from the implication about your competitors than the false claim in general.

OpenAI is against this bill, as written in the article.

Here in Argentina had a sunfloer oil brand that claimed to be cholesterol free (never claimed to be the only one). Anyway they were forced to add a small remark that says "like all vegetable oils".

A bit like non-GMO, gluten free salt.

I feel like other comments in this thread are not quite getting what the story is saying.

OpenAI (and probably others) do not like the bill. Their employees do on account of (possibly among other things) the whistleblower protections offered.

If you're going to write the same thing regardless of what the story actually says (e.g. regulatory capture is bad) what's the point of writing it?

please regulate our competitors

The HN title is misleading and clickbait. OpenAI is not supporting. Several employees of OpenAI et al may do.

The title is "OpenAI, Anthropic, Google employees support California AI bill"

That's the HTML <title>.

The actual headline is more accurate: "Exclusive: Employees at top AI firms back California's AI bill"

That's the first step to regulate artificial intelligence or LLM market.

Enjoy free and open model like lama or mistral while it lasts because openai and Google has way enough money to lobby politicians.

Personally I find that the idea of LLM being dangerous is laughable, there is nothing llm are going to give you that Google or bing can't already.

That's just 101 on killing entrepreneurship and competition in a free market, regulate it until only the big one can be part of it.

Four of us is enough

Reply on news.ycombinator.com