Ryan Shea ✓

@ryaneshea · USA · joined 02 Mar 2010

AI+bio experiments: https://t.co/JbiVsGf0Qi. Fmr Senior Advisor @US_FDA. 10+ unicorn investor. Cofounder @Stacks. MechE/CS @Princeton. NJ Math League State Champ.

30 178Followers

3 831Following

12 826Posts total

1MViews on collected posts

Latest posts

Ryan Shea ✓@ryaneshea16 Sep

Reminder that OpenRouter token consumption is not even remotely representative of token use across the industry. 1. Historically, OpenRouter has had smaller shares of OpenAI and Anthropic tokens because it was the router that engineers would use for the long tail of requests but 994 views · 8 likes · 2 reposts · 5 replies Open on X →

Ryan Shea ✓@ryaneshea14 Sep

Wait, so Pantheon was prescient after all? Artificial Intelligence requires obscene amounts of compute to discover effective but inefficient networks. Meanwhile, Uploaded Intelligence let's us take systems from nature that vastly surpass AI's on a per byte and per watt basis. 275K views · 4.2K likes · 228 reposts · 85 replies Open on X →

Ryan Shea ✓@ryaneshea8 Sep

https://t.co/cQAPPPzVrm

1.4K views · 8 likes · 0 reposts · 0 replies Open on X →

Ryan Shea ✓@ryaneshea8 Sep

Astra estimates the probability of a Millennium Prize problem being solved in 2026 as 0.7% https://t.co/9hCL76QVML

37.3K views · 692 likes · 16 reposts · 26 replies Open on X →

Ryan Shea ✓@ryaneshea4 Aug

As you can see below, Kimi K3 is one of the main models Bitcoiners are using to patch their software and improve industry security. GLM-5.2 was the model that HuggingFace used to patch its software when it was being hacked. Bitcoiners and HuggingFace are not using OpenAI and 10.6K views · 117 likes · 19 reposts · 8 replies Open on X →

Ryan Shea ✓@ryaneshea4 Aug

red teaming bitcoin: - we’ve written multiple harnesses and we’re launching a huge wave of reviews against many core bitcoin projects: crypto libs, wallets, infra, … - situation is extremely bad. - we’re averaging on the order of 1 critical exploit per hour per person. - we’ve 290.6K views · 2.4K likes · 328 reposts · 122 replies Open on X →

Ryan Shea ✓@ryaneshea12 May

Today I’m launching AI IQ — frontier AI models, scored on the human IQ scale. Instead of endless leaderboard tables, AI IQ shows: • Where models land on the IQ bell curve • How frontier IQ is changing over time • How models compare on IQ and EQ • What intelligence costs in http

425.3K views · 1.5K likes · 223 reposts · 162 replies Open on X →

Against accounts of the same size

6 posts from the last 90 days, next to the 10K–100K follower range. shown to more people than peers of the same size.

Median views23 932this account924median for 10K–100K

Reach, %79.30%this account3.62%median for 10K–100K

Engagement, %1.46%this account1.52%median for 10K–100K

Metric	This account	Median for 10K–100K	Ratio
Median views per post	23 932	924	25.9×
Reach (views ÷ followers)	79.30%	3.62%	21.9×
Engagement rate	1.46%	1.52%	0.96×

Others in this range → Compare with another account → How these benchmarks are built →

Growth & engagement

How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.

Views per post

425.3K12 May

290.6K4 Aug

10.6K

37.3K8 Sep

1.4K

275K14 Sep

99416 Sep

Last 7 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

0.46%12 May

1.02%4 Aug

1.40%

1.99%8 Sep

0.57%

1.64%14 Sep

1.51%16 Sep

Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.52%.

What the audience does

Likes66.8%8 905 in total

Reposts6.1%816 in total

Replies3.1%408 in total

Quotes1.8%240 in total

Bookmarks22.2%2 965 in total

Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.

The follower curve appears once this account has two daily snapshots — we take one a day, and this one is on its first.