Twitxy Saved

Bayesian

Verified account This account is verified. Learn more

@Bayesian0_0

#1 AI forecaster on Manifold Markets (and #2 across all categories)
manifold.markets/Bayesian
I want everything to make sense
Joined July 2024
2.9K Following 1.6K Followers 1.5K Posts

Latest posts 5

Bayesian Verified account @Bayesian0_0
You’re not talking to someone who woke up a loser. That loser attitude, that loser premise makes no sense to me.
Chris McGuire Verified account @ChrisRMcGuire 2d
The story keeps getting crazier. To recap everything that has been reported so far:

- The US intelligence community discovered that Jensen Huang, the CEO of the largest company in the world, was sleeping with Alice Huang, a Chinese spy and an executive at Megaspeed. (FT)
- Multiple independent investigations have concluded that Megaspeed is likely a Chinese cutout, which helps Chinese companies access AI chips or circumvent export controls. (NYT and Bloomberg)
- The US intelligence community also concluded that Alice Huang’s goal was to convince Jensen to lobby the U.S. government to loosen export controls on advanced AI chips. (NYT)
- The intelligence community relayed this information to “a small group of top US government officials” earlier this year. (FT)
- Over the last year, Jensen Huang has repeatedly lobbied the U.S. government to loosen export controls on advanced chips. @BISgov has done so, not only by permitting the export of H200 chips to China, but also not enforcing export controls that would have required a license for exports of the most advanced Blackwell chips to Megaspeed and Southeast Asia.
- Although @CommerceGov leadership was presumably briefed on this information by the U.S. intelligence community, @BISgov has taken no actions to tighten controls on Megaspeed or Southeast Asia.
- Megaspeed remains a preferred cloud partner of Nvidia, and continues to receive Nvidia Blackwell chips despite employing a Chinese spy who reportedly is romantically engaged with the CEO of Nvidia.
- Tens of billions of dollars worth of AI chips have been exported to Southeast Asia this year alone, and Malaysia and Singapore are now two of the four largest importers of AI chips in the world.

What is going on here??? We need answers. The question of who gets the most advanced AI chips in the world is literally the defining strategic question of our time. And not only does it seem like the MSS may be directly influencing the debate on export controls in the United States, but also @CommerceGov is refusing to listen to the warnings of the IC on this issue. What is a POSSIBLE good explanation for why Megaspeed is still receiving chips? Why is there not a license requirement for AI chip shipments to all of Southeast Asia? This is beyond strange at this point.

The fallout from this story will be massive. Obviously, the motivations underpinning Nvidia’s policy views opposing export controls must now all be questioned. But also, how did @BISgov sit on this information and has still taken no actions? We need answers.
View quoted post on X
2 replies 0 reposts 36 likes 922 views View on X
Bayesian Verified account @Bayesian0_0
I'd advocate for these kinds of visualizations of logistic accuracy vs capability index to benchmark models they know would be around the floor of the benchmark. So for example, for benchmark 5, testing 120 AECI models and observing whether they obtain the roughly 1% score that is predicted. This would give some information about the slope of the curve that the handful of bunched up datapoints cannot
2 replies 0 reposts 25 likes 716 views View on X
Bayesian Verified account @Bayesian0_0
Imo the pre-2026 slower trend has been mostly caused by a global capability trend at that time being faster than anthropic’s capability trend (especially openai improving more quickly in terms of benchmark perf); the mythos preview speedup was largely a trendbreak in terms of scale; and the subsequent speedup was largely a trendbreak in ai-led ai r&d acceleration, that will continue into 2027. I think this acceleration will not continue accelerating (lol) as much as many are saying; eg I am v skeptical of some kind of software-only singularity resulting from this kind of iterative acceleration in experiment efficiency
METR @METR_Evals 4d
1. Benchmarks provide one piece of evidence for acceleration. Since Oct 2025, AECI scores for Anthropic’s models have grown at 1.9x [95%: 1.5–2.8] the earlier rate. Analysis choices affect this estimate, and it remains unclear how benchmark scores map to real-world capabilities.
View quoted post on X
4 replies 2 reposts 56 likes 4.9K views View on X
Bayesian Verified account @Bayesian0_0
still working out the differences between the Bayesian ECI and Bayesian's ECI but I'm bullish! One of the lessons I've learned having a bunch of iterations of trying to find improvements to the original ECI methodology with claude is that the very weak models move around a lot. Ig i kind of think that from now on the anchors should no longer be 3.5 Sonnet at 130 and GPT-5 at 150 but instead be some kind of aligning of the axes with the original eci so that we can have comparable numbers across methodological iterations, ie shifting the slope is kind of just reinterpreting every number relative to that y-axis stretch. BECI stretches its axis to match the ECI as best as it can instead of using the 2 model anchors, for this reason.
Epoch AI Verified account @EpochAIResearch 4d
We’ve updated the Epoch Capabilities Index (ECI) to use a Bayesian methodology. This improves uncertainty estimates and lets us compare models more accurately, even when their individual ECI uncertainty intervals overlap.

Frontier model scores are largely unchanged: models stronger than GPT-4 move by less than 1 ECI point on average.
View quoted post on X
3 replies 0 reposts 21 likes 766 views View on X
Bayesian Verified account @Bayesian0_0
the Bayesian methodology ECI, not to be confused with Bayesian's ECI-like methodology
Epoch AI Verified account @EpochAIResearch 4d
We’ve updated the Epoch Capabilities Index (ECI) to use a Bayesian methodology. This improves uncertainty estimates and lets us compare models more accurately, even when their individual ECI uncertainty intervals overlap.

Frontier model scores are largely unchanged: models stronger than GPT-4 move by less than 1 ECI point on average.
View quoted post on X