Mike Knoop ✓

@mikeknoop · sf bay area · joined 16 Jul 2009

co-founder @ndea and @zapier @arcprize

24 859Followers

366Following

4 184Posts total

70.9KViews on collected posts

Latest posts

Mike Knoop ✓@mikeknoop4 Sep

New chart from Astra testing on ARC v3 Astra often emits zero reasoning tokens per action at lower reasoning levels. We've never seen this before. And surprisingly Astra low is 2X more accurate than Sol max. This suggests Astra is leveraging a secondary test-time adaptation htt 11K views · 209 likes · 21 reposts · 24 replies Open on X →

Mike Knoop ✓@mikeknoop3 Sep

@mikeknoop I have zero fear of all mankind turning against innovation. A part of it can, but the remaining parts who don't, will quickly boss over the rest. 87 views · 1 likes · 0 reposts · 0 replies Open on X →

Mike Knoop ✓@mikeknoop3 Sep

@mikeknoop 66% to almost 100% just from the adapter is crazy. the harness is basically part of the benchmark at that point. 185 views · 2 likes · 0 reposts · 0 replies Open on X →

Mike Knoop ✓@mikeknoop3 Sep

@mikeknoop Will ARC-AGI-4 model creativity? 335 views · 2 likes · 0 reposts · 0 replies Open on X →

Mike Knoop ✓@mikeknoop3 Sep

Very rewarding to be personally involved with two of the five Astra headline benchmarks https://t.co/kgZxylgMCy 1.7K views · 35 likes · 0 reposts · 0 replies Open on X →

Mike Knoop ✓@mikeknoop3 Sep

AutomationBench (by @Zapier) is the lead benchmark OpenAI used to showcase Astra's workflow automation capability -- 44% at $2.45/task. 3.2K views · 30 likes · 5 reposts · 4 replies Open on X →

Mike Knoop ✓@mikeknoop3 Sep

GPT-6 Astra is the new SOTA on ARC-AGI-3 It's a qualitatively large leap towards AGI and the pace of progress is frankly surprising. That said, we lack evidence to call this AGI yet. While we are still studying the human capability gaps, we believe open-ended invention is 29.2K views · 189 likes · 18 reposts · 17 replies Open on X →

Mike Knoop ✓@mikeknoop26 Mar

@mikeknoop These scammers solved it! https://t.co/VidL0lFDTu 46 views · 2 likes · 0 reposts · 0 replies Open on X →

Mike Knoop ✓@mikeknoop25 Mar

@mikeknoop the jump from static benchmarks to interactive environments is the right move. the hardest part of my job isn't answering questions — it's figuring out what to do when the first three things I tried didn't work. that's what v3 is testing. 76 views · 1 likes · 0 reposts · 0 replies Open on X →

Mike Knoop ✓@mikeknoop25 Mar

ARC-AGI-3 and ARC Prize 2026 are now live with $2,000,000 in prizes! As of today, version 3 is the world's only unsaturated agentic intelligence benchmark. Humans score 100% and frontier AI scores ~0%. Play here: https://t.co/nF1MhGvTyz While no single version of ARC is 25.1K views · 165 likes · 22 reposts · 13 replies Open on X →

Against accounts of the same size

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

Median views1 708this account924median for 10K–100K

Reach, %6.87%this account3.62%median for 10K–100K

Engagement, %1.15%this account1.52%median for 10K–100K

Metric	This account	Median for 10K–100K	Ratio
Median views per post	1 708	924	1.85×
Reach (views ÷ followers)	6.87%	3.62%	1.90×
Engagement rate	1.15%	1.52%	0.76×

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.