Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

David Sacks: OpenAI and Anthropic Don't Need Regulations to Pace Frontier Models

twitter.com

Post reply on HN
Print view View original

221–230 of 267 posts

Re: David Sacks: OpenAI and Anthropic Don't Need Regulations to Pace Frontier Models

#221

Post by 0xDEAFBEAD »

bigyabai wrote: post #216
0xDEAFBEAD wrote: post #159

Earlier quoted context omitted.

With closed-weight, we can get the models out to defenders before they're in the hands of attackers. There are also other categories of risk such as biorisk. Open-weight means no guardrails. Don't get me wrong, I like your idea of deflating the AI bubble, I'm just skeptical that this is a great way to do it. The PauseAI people have some ideas: https://pauseai.info/proposal * International body which must approve new…

> Open-weight means no guardrails. We're going to live in that world forever now, capable adversarial models have already been trained and will be used in a large conflict. The pragmatic issue is hacking, and current open-weight LLMs are already capable of it. Bioweapons and RSI are pie-in-the-sky theories that distract from the real-world situation that each government wants to exploit.

>Bioweapons and RSI are pie-in-the-sky theories that distract from the real-world situation

It wasn't long ago when automated hacking was a pie-in-the-sky theory.

Sometimes it's helpful to prepare for things in advance. An ounce of prevention can be worth a pound of cure.

Re: David Sacks: OpenAI and Anthropic Don't Need Regulations to Pace Frontier Models

#222

Post by gundmc »

SpicyLemonZest wrote: post #212
dehrmann wrote: post #144

Earlier quoted context omitted.

> you will face criminal and financial penalties for damages caused Even with where we're at right now, if people involved with the HF or Ruby Gems attacks faced legal consequences, the industry would get a lot more conservative very quickly. And unlike social media, section 230 doesn't apply, and there aren't any free speech concerns.

The HF and Ruby Gems attacks came from safety testing, where the AI providers attempted to understand what negative consequences could arise if someone subverted their safety protections. Don't you worry that legal consequences would disfavor safety testing and thus create more rather than fewer problems?

I thought it came from RLVR with poor sandboxing and not explicitly security testing.

Re: David Sacks: OpenAI and Anthropic Don't Need Regulations to Pace Frontier Models

#223

Post by byzantinegene »

hrpnk wrote: post #41

I wished OpenAI and Anthropic IPOd already. They just try to outcompete each other by finding the most splashy headlines undercutting prices in unhealthy and unsustainable means long-term.

they would if they could, their investors are seething already.

Re: David Sacks: OpenAI and Anthropic Don't Need Regulations to Pace Frontier Models

#224

Post by bigyabai »

0xDEAFBEAD wrote: post #221
bigyabai wrote: post #216

Earlier quoted context omitted.

> Open-weight means no guardrails. We're going to live in that world forever now, capable adversarial models have already been trained and will be used in a large conflict. The pragmatic issue is hacking, and current open-weight LLMs are already capable of it. Bioweapons and RSI are pie-in-the-sky theories that distract from the real-world situation that each government wants to exploit.

>Bioweapons and RSI are pie-in-the-sky theories that distract from the real-world situation It wasn't long ago when automated hacking was a pie-in-the-sky theory. Sometimes it's helpful to prepare for things in advance. An ounce of prevention can be worth a pound of cure.

But hacking has always been a vulnerability, even pre-AI. The scale multiplier here is the number of vulnerable machines, not the intelligence of AI.

> An ounce of prevention can be worth a pound of cure.

It can be, but if you get chemotherapy on a hunch instead of a diagnosis then it could kill you instead. There is no evidence that RSI will ever exist.

Re: David Sacks: OpenAI and Anthropic Don't Need Regulations to Pace Frontier Models

#225

Post by dmode »

pembrook wrote: post #130
okdood64 wrote: post #104

Earlier quoted context omitted.

> his takes on the Ukraine war What's specifically?

Mainstream narratives for the first year or two of the war were: - Russia was a paper tiger and weak - Ukraine was going to win against a corrupt/unmotivated Russian military so it should keep fighting - Putin had made an error and he would get overthrown - The Russian economy was going to collapse due to sanctions - Russia oil pipelines would freeze and never come back online without western help (this was a Peter Z…

What? Everyone I know thought Russia would absolutely destroy Ukraine within days and take it over

Re: David Sacks: OpenAI and Anthropic Don't Need Regulations to Pace Frontier Models

#226

Post by hn_throwaway_99 »

jamienk wrote: post #213
hintymad wrote: post #171

Earlier quoted context omitted.

> Right now, overall compute is a moat. Very true. Or further, access to capital is the moat. It is the very reason that we don't have a real open-source community that trains frontier models - individuals simply can't afford the training infrastructure, nor sufficient high-quality training data.

https://hugovergnes.github.io/little-lm-3-8b/ Remember how DeepSeek v.whatever cost ~$5m Stable Diffusion 1.5 was reportedly $70k in compute.

None of those are frontier models, nor could they have been. E.g. DeepSeek's "$5 million" included piggybacking off OpenAI/Anthropic's hundreds of millions/billions by using distillation.

Re: David Sacks: OpenAI and Anthropic Don't Need Regulations to Pace Frontier Models

#227

Post by rexpop »

gr_norm wrote: post #98
matheusmoreira wrote: post #66

Earlier quoted context omitted.

Couldn't agree more. Better yet: the AI labs get straight up nationalized.

No. That kills competition. I want everyone to be competing on the open playing field they are right now, with losses socialized as minimally as possible. There's nothing fundamentally special about LLM companies that would warrant nationalizing them.

To what degree are the competing Chinese labs nationalized?

And—although this is a different point—it's their flagrant exploitation of information commons that warrants their special treatment.

Re: David Sacks: OpenAI and Anthropic Don't Need Regulations to Pace Frontier Models

#228

Post by frm88 »

deagle50 wrote: post #56

I'll go a step further, they aren't just pretending, they are blackmailing the country. The sane response would be "we don't negotiate with terrorists, and you will face criminal and financial penalties for damages caused."

you will face criminal and financial penalties for damages caused.

No you won't. The US president, yesterday:

U.S. President Donald Trump on Sunday likened critics of artificial intelligence to "very negative forces" bringing up scenarios that will not happen, and said he wanted to make sure that the U.S. remains the industry leader. "We're leading China in AI. We're the most sophisticated country in the world, and frankly I want to keep it that way because whoever wins AI wins," Trump told reporters at his Ireland golf course, when asked if the AI industry should slow down or be more regulated.

https://www.reuters.com/world/europe/trump-says-very-negativ...

Re: David Sacks: OpenAI and Anthropic Don't Need Regulations to Pace Frontier Models

#229

Post by frm88 »

SpicyLemonZest wrote: post #219
filearts wrote: post #218

Earlier quoted context omitted.

Let's say that we are "pacing the frontier", and an approved model goes and causes billions of damage, or worse, loss of life. The auditor signed off on the model, so who is to blame? Who put their akin in the game? METR did. The public did. But it seems like the lab has crafted a system in which they remain blameless. The calculus of risk is fundamentally different when the adverse consequences are obvious and tangi…

Again, you're presenting ideas that I just don't understand where they're coming from. Why would the lab be blameless in such a situation? You and I and METR and the public would all want the frontier labs to face adverse consequences if their products caused huge problems, so who would give them immunity and why?

You and I and METR and the public would all want the frontier labs to face adverse consequences if their products caused huge problems

That makes you very negative forces, according to the US president:

https://www.reuters.com/world/europe/trump-says-very-negativ...

Re: David Sacks: OpenAI and Anthropic Don't Need Regulations to Pace Frontier Models

#230

Post by socalgal2 »

mrweasel wrote: post #21
socalgal2 wrote: post #4

Admits China won’t stop but says USA’s top should. Reminds me of City in the Edge of Forever

Unsure about your reference, but is your argument: Someone else is doing it, so I should be allowed to as well. That's basically the western companies complaining that they can't compete with China because Chinese companies are allowed to polute more, so they would be allowed to polute as well. I don't buy that. Once you realise that you're doing something wrong or dangerous, you stop, regardless of what everyone els…

the reference is that the idea is good but at the wrong time can lead to disastrous outcomes.

Some people say AGI/ASI will destroy the world. Others say, whoever makes it first gets to rule the world. So if slowing means China gets it first and it doesn't destroy the world then you'll be living under their thumb by slowing down now.

Post reply on HN