@ansonwhho

Researcher @EpochAIResearch

 Toronto, Canada
ansonwhho.github.io
Joined January 2022
  • Tweets64
  • Following1.1K
  • Followers1.1K
6Media
How do Chinese AI firms make money? In a new report, @cherylwoooo and @ansonwhho identify five revenue sources and analyze how much each contributes. 🧵
13
86
28
439
83,074
Really glad this is out! Came as a big shock to me when my colleagues pointed out this possibility
Epoch AI
@EpochAIResearchAug 24
We discovered that US GDP statistics miss most of the value Nvidia adds to the US economy. As a result, GDP growth has been understated by ~0.3 percentage points over the last year.
1
37
1,902
Introducing FrontierMath Tier 4: a benchmark of extremely challenging research-level math problems, designed to test the limits of AI’s reasoning capabilities.
17
61
24
557
88,485
I strongly disagree with @lugaricano’s thread on explosive growth. While the thread raises important points, I think it fails to get to the core of the reasons to believe that explosive growth is plausible – allow me to explain!
Luis Garicano 🇪🇺🇺🇦
@lugaricano28 Jun 2025
1/@EpochAIResearch doubles down on preiction AI will drive 20%+ annual GDP growth. Economists remain skeptical. This is the defining debate of today: AI builders see infinite prosperity ahead. Economists see the same limits that constrained every technological revolution.🧵 1/13
2
3
1
59
8,293
more replies
I do agree that often "progress works itself out of a job", baumol effects and human bottlenecks are important and can make explosive growth less likely (especially in the next few years)
1
7
834
But I think that research on explosive growth has already provided strong counterarguments to the points that @lugaricano raises
7
732
How do reasoning models solve hard math problems? We asked 14 mathematicians to review o3-mini-high’s raw, unsummarized reasoning traces on 29 FrontierMath problems. Here’s what they found:
17
89
26
523
110,411
Our AI Supercomputers database is now live! Explore specs for 700+ training & inference clusters—compute, chip types, power draw, ownership, location, and more.
4
50
16
279
49,093
Is GPQA Diamond tapped out? Recent top scores have clustered around 83%. Could the other 17% of questions be flawed? In this week’s Gradient Update, @GregHBurnham digs into this popular benchmark. His conclusion: reports of its demise are probably premature.
5
21
5
235
49,016
Is AI already superhuman at FrontierMath? To answer this question, we ran a competition at MIT, pitting eight teams of mathematicians against o4-mini-medium. Result: o4-mini beat all but two teams. And while AIs aren't yet clearly superhuman, they probably will be soon.
18
79
17
468
117,359
How likely is an intelligence explosion as forecast in AI 2027? Algorithmic advances that could drive an intelligence explosion may be bottlenecked by compute, according to new research from @noshpesoj and @uchicagoxlab described in this week’s Gradient Update. Here’s why:
9
30
17
262
48,449
Where’s my ten-minute AGI? If AI systems are able to perform tasks at 1-hour time horizons, as @METR_Evals found, why hasn’t AI automated more tasks already? In a new Gradient Update, @ansonwhho argues this is due to three fundamental reasons.🧵
6
43
3
245
24,379
Why have AI benchmark scores often felt disconnected from real-world usefulness? It wasn't just technical limitations to creating realistic benchmarks. Historically, predicting the real-world capabilities of AI systems simply wasn’t the main goal.🧵
11
32
8
278
71,531
Load more