Second Thoughts
Subscribe
Sign in
Home
First Takes
Archive
About
Latest
Top
Discussions
Anthropic’s Scion: A Four Year Project in Five Months
Current AI isn’t superhumanly intelligent, but it is superhumanly fast, and that opens many new possibilities
Oct 14
•
Steve Newman
30
4
2
What I saw around the Curve, 2026
The vibes, they are a-changin’
Oct 12
57
9
7
September 2026
14 Reasons Robotics is Hard
And why you should ignore demo videos
Sep 1
•
Steve Newman
63
12
10
August 2026
Cybersecurity: Let’s Play to Win
Stop expecting perfection and start making it unnecessary
Aug 21
•
Steve Newman
24
5
2
July 2026
Anecdotes Everywhere, Evidence Almost Nowhere
The state of AI in mid 2026, part 3
Jul 20
•
Steve Newman
29
4
June 2026
Never Trust A Number
They're shortcuts to understanding, and there are no shortcuts
Jun 30
•
Steve Newman
23
2
May 2026
AI Can Help Plan a Bioweapon. Building One is Still Hard.
Hurdles include facilities, materials, and steps requiring unwritten lore and hands-on experience.
May 28
•
Abi Olvera
24
4
2
From Compute Overhang to Compute Crunch
The state of AI in Q2 2026, part 2
May 19
•
Steve Newman
33
13
4
Is AI 2027 Coming True?
The state of AI in Q2 2026, part 1
May 7
•
Steve Newman
133
25
13
April 2026
Tacit Knowledge: The Missing Factor in AI Bio Risk Assessments
Lab skills come from hands-on mentorship, not from reading the entire internet.
Apr 23
•
Abi Olvera
63
26
9
The Centaur Era
The question isn't what AI can do, it's what you can do with AI
Apr 17
•
Steve Newman
39
5
4
To Forecast AI's Impact on Biosecurity, We Asked: Why are Attacks So Rare?
Nine factors that practitioners say make bioweapons rare.
Apr 8
•
Abi Olvera
45
11
4
This site requires JavaScript to run correctly. Please
turn on JavaScript
or unblock scripts