Researcher on statistics and ML working with @bilbyai My work is on prompt tuning, ML modeling, multiple testing, resampling methods, random field theory.
Are you at #OHBM2024? Come check out my talk on Localized Cluster Enhancement - TFCE revisited with valid error control, in the oral session on advanced statistical methods on Wednesday at 11.30. Or poster 1871, on Wednesday/Thursday to see how to improve the validity of TFCE (1)
The best thing about AI content creation is that it doesn't take nearly as much time as people think.
I'll come up with an idea over breakfast, talk it through with my AI assistant, set it rendering, and get on with my day job.
At lunch, I'll review the result, give some feedback, and send it off again. By dinner, there's another revision waiting.
I'll spend the evening playing with the kids, then check the latest version before bed. If it needs work, I'll give it more instructions and have another version ready in the morning.
The beauty of AI is that it keeps working while you're doing other things. You're only spending small pockets of time directing and refining, rather than hours actually producing the content.
@RealAlbanianPat right but couldn't they pay for a single subscription and use that? why would it run to hundreds of dollars? Or do you mean multiple different people using their accounts to do scraping?
Fable 5.2 and Mythos 5.2 available now, some thoughts:
(1) if Anthropic is going to have super powerful internal models I generally think they should try hard to release them publicly soon after for power concentration/transparency reasons (with appropriate guardrails on areas where they are creating new risks).
(2) After the METR report/HF/various sober minded people I know increasingly changing their minds on timelines and freaking out more etc I am finding the speed of AI progress viscerally scary and new model releases are not fun and exciting for me like they used to be.
(3) The situation is complicated, but on an emotional level I deeply resent AI companies like Anthropic for threatening to impose massive risks on me and everyone else without making it a top priority to ameliorate those risks in the same way it is a top priority to hit new revenue targets and train shiny new models. I know a lot of folks there want to do better but they are engaging in extremely perilous actions, are benefitting tremendously in money power and status, and need to be held to a high standard that they are not meeting.
(4) I'm gonna use these models and they will probably be useful and cool and stuff.
@Autonomous_Chad@whalepoacher1 They are cherry picked because you chose specific races and a metric to suit your narrative whereas lots out there shows the ither point e.g. the nyt article. Also it sounds like you're mixing house and senate data which is statistically dodgy.
@nkagan_official Also I don't think it makes sense to mix the house and senate and on the senate Kalashi was actually below the model average - way below the model leaders
@nkagan_official I don't lnow why you'd pick 9 am on election day. If anything it should be the opposite. I.e. its more impressive to predict further in advance.
The betting markets lagged the high quality models in forecasting a blue wave the ENTIRE cycle.
The way we judged forecasts at FiveThirtyEight was based on the accuracy of its output *each day* it was live. Not just the morning of the election.
Kalshi thinks you are dumb.
Dario — Thank you for taking the time to respond to my previous post. I think it's great when this kind of engagement happens in public rather than on company Slacks, and I hope to see more of it.
I agree that Anthropic isn't best modeled as a profit-maximizer. Your post gave several compelling examples of Anthropic taking competitive hits for safety, and insofar as my previous post suggested otherwise, that's on me. But I don't think I made my main point as clearly as I could have, so let me try again. It comes down to four key claims:
1. Anthropic seems to be racing at breakneck speed to build recursively self-improving superintelligence (RSI). By its own lights, that could kill billions of people or literally destroy the world.
2. Risking destroying the world is very hard to justify. It is only conceivably acceptable if serious attempts to put RSI under robust societal governance are failing.
3. Far from meeting that burden, Anthropic, when push comes to shove, often works to keep governments from tying its hands. It seems to care much more about preventing government overreach than about reducing the risk that it and other AI companies will destroy the world.
4. Even if the AEGIS Act really was an honest mix-up, as you suggest, that is itself a serious problem. It suggests Anthropic isn't treating the governance of this technology with anything close to the gravity it deserves.
In greater detail for those who want it:
1. Anthropic is racing to RSI, and by its own lights that is insanely risky.
Launching RSI could literally destroy the world. That's according to you, your co-founders, and most of your safety staff and capabilities researchers, not to mention many AI researchers outside your walls. You are intentionally trying to create generally intelligent systems more capable than all humans that you claim in six to twelve months “could be capable” of taking over the entire internet before we have the ability to reliably monitor, control, or align them.
“Succeeding” at RSI could kill or permanently disempower billions of people in the near future, including me and everyone I love. Yet Anthropic’s primary focus seems to be marshalling an immense amount of resources and talent in a race to get there first.
2. Risking destroying the world carries an enormous burden of justification.
I believe the vast majority of people do not want Anthropic to do this (to the extent they know what it's doing). Imposing even a modest chance of catastrophe on billions of people who never consented requires an extraordinary justification. The central one Anthropic has offered: someone will build RSI regardless, so it's better for a safety-focused lab to be at the front of the pack, using that position to support strong AI governance. (Other reasons Anthropic has given, like modestly accelerating improvements in healthcare or economic growth, don’t seem remotely proportionate to the risks Anthropic is imposing by its own lights.)
But that argument only works as a last resort. Racing to RSI is only conceivably okay if attempts to institute robust societal governance to manage that race are failing despite your best efforts. A lab that relies on this justification should be a strong, consistent supporter of efforts to enact governance that could bind it, including governance that could slow it and significantly constrain what it does.
3. Rather than supporting efforts at strong regulation, Anthropic has intervened to keep governments from tying its hands.
Anthropic has sometimes helped: it played a very helpful role in passing transparency bills in Illinois and ongoing efforts in Massachusetts, and I'm sincerely appreciative. But transparency is far from enough for a technology that will be the world's most powerful superweapon at best, and an uncontrollable swarm of lethally superintelligent agents at worst. When it matters most, Anthropic has often pushed for weak regulation.
The AEGIS Act is the most obvious example. No highly risky industry is regulated as anemically as AEGIS would regulate frontier AI: not nuclear, not banking, not aviation, not food safety, not the manufacture of explosives. In all of those industries, specialist regulators can address unsafe practices well before they cause harm. AEGIS would bar the federal government from intervening to block unsafe development or deployment of models until there is an "imminent risk" of catastrophic harm, and would require it to meet a demanding burden of proof in court. Worse, AEGIS would also prevent states from picking up the slack. This design makes perfect sense if you're a company trying to minimize the risk that the government can overly constrain you. It makes little sense if you want American democracy to govern frontier AI effectively.
This isn't the first time. The initial version of California's SB 1047 would have let the Attorney General sue to stop developers like Anthropic from imposing unreasonable risks of catastrophic harm. It would also have created a regulatory agency to help set safety standards. Anthropic objected, and helped get both provisions removed before ultimately offering some support for the bill (which I also genuinely appreciate). I’m not saying that Anthropic should have supported 1047 in its initial form; that’s a complex issue and I realize there were questions about whether earlier versions of the bill were less politically feasible. Still, in Anthropic's own words, it helped weaken the bill because "[t]here is a substantial risk that the bill and state agencies will simply be wrong about what is actually effective in preventing catastrophic risk, leading to ineffective and/or burdensome compliance requirements."
The same goes for New York's RAISE Act, whose initial version also would have let the Attorney General sue to stop unreasonable risks of catastrophic harm. Anthropic credits itself as a supporter of RAISE, but it did not support the bill while it was in the legislature, including when it nearly died. After the bill passed the legislature, Anthropic publicly critiqued it, saying "Any state proposals should be narrowly focused on transparency and not overly prescriptive."
Why does Anthropic consistently take the risk of clumsy government action far more seriously than its own repeated warnings that companies like Anthropic will literally destroy the world by racing to RSI?
4. Even if AEGIS was an honest mix-up, that is itself a serious problem.
Your response frames AEGIS as an early draft and a miscommunication. Perhaps you would have strengthened it and narrowed its preemption. (That wouldn't explain Anthropic's interventions to weaken SB 1047 or RAISE, but leave that aside.)
Suppose that's exactly right. A federal framework for governing RSI is one of the highest-stakes documents your company will touch. Your post admits Anthropic "could have handled [these matters] better," but its tone ("This kind of thing happens a lot in large organizations") doesn't reflect the gravity of this issue. Anthropic talks about the need for “operational excellence” in critical areas, and I have a hard time believing that a similarly cavalier attitude would be tolerated while Anthropic was negotiating a key compute deal or preparing its next large pre-training run.
In short
The justification for racing rests on Anthropic fighting for strong governance, and Anthropic's record on regulation points the other way. Together with its breakneck racing, that fits a much darker picture of the company than the one that seems to prevail among your employees. I hope that Anthropic reflects on this picture, supports stronger regulation, exercises extraordinary caution when advocating for weak regulation, and doubles down on its founding purpose of helping the world safely navigate transformative AI together, instead of prioritizing Anthropic’s ability to win a race that imposes intolerable risks on everyone alive.
Everyone hates on Gemini. But it consistently wins as the best model in terms of cost efficiency across a range of our different internal experiments. Best must also factor in cost and I think many forget to take that into account.
It’s genuinely astonishing how bad Gemini is.
Google has invested billions in AI, only to produce a product that is basically unusable.
And somehow, they’ve managed to make Search worse along the way.
Impressive, really.
@julianharris Erm no clef itself is 10 times jev prices so not a competitor. Clef flash is over double the cost and there isn't really evidence that its better (e.g. it only has 4-3 clear wins above). The whole point of Jev is that it's cheap so I'm not sure what that proves.