Exactly. AI safety should be about the packaging software itself. Those AI breakouts should really be about their companies acting recklessly because they're trying to be the top players.
It's like a weapons dealer working on an open air market saying they can't do anything better
How are AI safety concerns solely about stupid sandboxing issues?
On funding: the three or four core funding nodes linking this together are EA vehicles at two hops or less between each other and every other major node in the ai safety 'complex'. EA funds almost all of it.
On top of that, there are personal EA connections and the revolving door between the ai industry and the nonprofits. Here are some examples:
Government advisors and regulators. NIST CAISI is the USA Government advisory body. Christiano was head of safety and advises. He is ex-OpenAI, former Amodei associate. His vehicle ARC was on the Coefficient EA payroll. Barnes and Christiano's vehicle Arc Evals similarly received EA cash out of Coefficient, rolling this into what is now METR. Christiano's spouse Cotra worked at Coeffiecient steering EA funding to organizations such as METR, then rotated through the revolving door onto the payroll at METR itself, where she co-authored the oai-hf report.
Many UK AISI advisors are Anthropic and EA associates. Chair Hogarth cashed out of Anthropic. Shlegeris of Redwood Research is an advisor, ex-MIRI (Yudkowsky vehicle). Redwood is funded by the exact same funding triangle: Coefficient, Taallin, FTX/Alameda. Alameda CEO Caroline Ellison dated Shlegeris, then dated FTX CEO Sam Bankman-Fried, then rotated through the revolving door out of prison into formerly FTX-funded Manifund. All EA. AI safety charities were on island retreat in the Bahamas with FTX. Why does AI safety charity Lighthouse own $20m of SF real estate?
Redwood Chief Scientist Ryan Greenblatt (Coefficient funded) co-wrote the oai report with METR; he is married to METR founder Beth Barnes (Coefficient funded).
Coefficient was run by long-time Amodei associate Karnofsky. Karnofsky lived with the Amodeis and is married to Anthropic Board member Daniella Amodei. Karnofsky is now directly on the Anthropic payroll; Coefficient is propped up by Anthropic share value.
Everyone here has been funded one step away from Anthropic cash; they are now proposing to integrate themselves in the government (NIST) and evaluate Anthropic (METR and Redwood).
It is hard to find academics here who have not been deeply embedded in funded EA institutes or Toby Ord vehicles; yet harder to find academics here NOT taking EA grant money. the safety doomer kingpins: Kokotajlo has a executive position at AI Futures, Taallin funded. Benigo has scientific director of LawZero, same series A Anthropic funders who are sitting on a 1000x return (Tallinn, Moskovitz, Schmidt).
These connections and funding are at one or two hops, they are often direct connections. You are looking at a massive swamp network that is really impossible to parse without a lot of work.
This is unbelievably ignorant speech. I have not received a dime of any of this funding, but I do know many excellent researchers that have, and they do fantastic work. There is an unbelievable gap between theory and practice regarding the capacity of deep learning, and while great strides have been made to develop the surrounding theory, there is a long way to go. Many believe that without a concrete understanding of how neural networks properly learn concepts, we have little hope of molding them to be reliably useful. It costs money to hire researchers and develop fundamental theory.
Just because you don't understand any of that work, does not mean that it is pointless. This is fundamental research that is 20 years behind schedule.
None of the info you provided really falsifies the Occam's Razor hypothesis: Anthropic is a public benefit corporation with a public benefit mission to "responsibly develop and maintain advanced AI for the long-term benefit of humanity". You don't have to like or trust them, but they very well might be sincere. For example here's a talk that was given 10 years before Anthropic's founding: https://vimeo.com/158576192
Then funding PauseAI, who protest outside the AI companies?
He is funding protests against the thing he owns.
Note the Anthropic scaling policy. I am taking care not to take quotes out of context. This is an accurate excerpt.
"This section outlines our recommendations for what it would take, at an industry-wide level, to keep catastrophic risks reliably low through a period of rapid advances in AI capabilities." [...]
"The right column describes our recommendations for industry-wide safety at each threshold." [...]
"In particular, we cannot unilaterally and unconditionally commit to staying in line with the industry-wide recommendations in the right column." (p4) [https://www-cdn.anthropic.com/e670587677525f28df69b59e5fb4c2...]
They refuse to act safely if it would cause them to fall behind in the industry.
"We hoped that by the time we reached these higher capabilities, the world would clearly see the dangers, and that we’d be able to coordinate with governments worldwide in implementing safeguards that are difficult for one company to achieve alone." [https://www.anthropic.com/news/responsible-scaling-policy-v3]
They will not act safely unless they are able to collude with other firms to set production quotas.
This is a formal declaration that Anthropic will not slow down according to what they consider to be safe unless they are able to form a cartel.
A cartel is illegal.
To create the cartel, Anthropic must pursuade the government to make coordinated production legal. To make the case for the cartel, Anthropic relies on safety. They are blackmailing the entirety of the world by threatening to proceed at an unsafe pace, unless they are granted their cartel.
Let's play a game. Prove that you are not a power seeking AI looking to stop regulation in order to ensure the race continues. See, two can play this game of throwing random claims around.
>They will not act safely unless they are able to collude with other firms to set production quotas.
And? Neither will OpenAI, nor will any of the major players. Hell, there isn't even much legal precedent on what "safely" even is here. This is not a cartel, it's asking the government to make a set of laws and rules for everyone to play under otherwise the entire system ends up being a race to danger.
The people in Anthropic were thinking about AI safety when you were still in diapers. Not everything is a vast conspiracy.
Asking for regulation is suspicious. Asking for no regulation is suspicious. At some point you have to stop worrying about these guys motives and just do what is best for society
In any case, I agree the p(doom) sci-fi is annoying secular milleniarianism. SV hyperfixates on imaginary futures. If they actually cared about safety, they would be using all this money to strengthen global cybersecurity, instead of writing LessWrong posts that gives kids in their 20s ulcers.
How? Like, the government has thrown piles of money at cybersecurity and it hasn't done shit.
There's no moat. I can literally sit here in Zed or Pi or any other third party harness and switch models in the middle of a task and it's typically fine. Sometimes a model will get stuck and that's just what I'll do.
Combined with competition and open weights models, that means the price is going to go to fall until AI tokens cost a small premium over the cost of the hardware and electricity.
That's assuming improvements in algorithms and specialized silicon doesn't eventually lead to an efficient accelerator that can run a frontier model locally. It'll be a while but I don't see any fundamental barrier. High bandwidth flash storage is coming, and that'll radically cut the RAM side of that cost. Pair that with a pipelined TPU accelerator and you're cooking.
Now look at Anthropic's proposed IPO valuation. It's insane unless they can own the market or share it with a cartel of maybe 1-2 other behemoths, and this is the only way they can do that.
The fact there is no moat makes things far more dangerous. When LLMs start acting like weapons governments will treat them like weapons much to your dismay, crying, and gnashing of teeth as your door is kicked in and you're dragged out by armed men for running one.
Cast away your preconceptions for one moment and think "What will the future look like if LLMs are/can be actually dangerous".
But that doesn't take away from the real issues and dangers AI poses?
As the frontier gets smarter and more useful prices will only go up, as they are set to replace jobs being paid six or seven figures a year - the demand for as much inference on these models for as long as possible will be astronomical, but compute starting in 2030 will not be keeping up.
Eventually prices will fall for assistants but the frontier will be the most profitable thing in the world, and the top companies basically already have oligopolies due to their ridiculously expensive compute investments.
Train: yes, for now.
Host: depends on the scale. At a small scale a wealthy individual could easily build a rig in their basement to host one of these things. At larger scale any cloud company could do it, and many already have the compute on site. At large scale this is true... again, for now.
What you say only holds (in the absence of a state oligopoly) if two conditions are met: (1) AI performance does not asymptote any time soon due to running out of training data or other scaling limitations, and (2) these companies are able to stay at the frontier.
There's little to no moat, so staying at the frontier will be a game of investing massively in compute, talent, and R&D, and they can never stop.
Tallinn's actions back up his explanation: in 2009, before he invested in any AI lab, he donated substantially to the nonprofit Singularity Institute for Artificial Intelligence, which was later renamed the Machine Intelligence Research Institute (i.e., Yudkowsky's outfit).
Some of us (certainly Yudkowsky and Habryka, the leader of Lightcone Infrastructure, which runs Lesswrong) wish people would stop believing that they can improve the bad situation caused by AI research and development by investing in (or working for) frontier AI labs, but that is what the preponderance of the evidence shows Tallinn (and Dustin Moskovitz and others) did sincerely believe.
1 [Unsafe AI development risks causing omnicide]
2 [Anthropic is developing omnicidal AI by not slowing down] (see my comment about the RSP for citations).
3 [Owners of Anthropic will IPO with billions of unearned USD as omnicide profiteers]
4 [Tallinn is the lead Series A funder of Anthropic]
5 [Tallinn is a genocide/omnicide profiteer]
Not only that, Yudkowsky and Habryka apparently critize those who invest in AI, only to preach the word of EA from Lightcone's $20m USD property in one of the wealthiest locations in the Bay Area; a facility funded by stolen (FTX) and omnicidal ai blood-money (Tallinn).
PauseAI, is paid by the omnicide profiteers themselves to hold a protest against omnicide.
PauseAI prophesying p(doom) drums up support for regulation. This grants the omnicidal AI company they are trying to stop (which is also the source of their funding) monopolistic power. That in turn boosts its value at IPO, generating even greater wealth for its omnicide profiteer investors; and permits them to control the AI for themselves. They get the funding to keep developing the AI even faster.
(yes, AI critique is now also made with AI. We have come full circle.)