Anthropic shipped its most agentic model yet, put a whole lab bench inside the AI, and read sentences off brain waves -- all in one day. And on that same day, five independent research teams landed the opposite message: point these agents at a long, messy, real-world job and they fail most of the time. We dig into what they're actually failing at (it's not the clicking or the coding -- it's judgment), why bigger models are sometimes worse at knowing when to quit, a cheap no-retraining fix that doubled that skill, and a 35-billion-parameter agent arguing the whole industry has been scaling the wrong thing. Plus the day's headlines: the hidden fingerprint found inside Claude Code, Sonnet 5's cheaper-per-word-pricier-per-job twist, the export ban that reversed in 18 days, and more.

Chapters

0:00 The Invisible Mark in the Date
1:14 The Headlines
7:54 Intro -- The Reliability Reality
9:05 The Reliability Wall
13:57 Knowing When to Quit
18:03 Scaling the Horizon, Not the Parameters
21:33 Wrap-Up
23:44 Outro — “Hold the Thread”

Links

The Invisible Mark in the Date -- https://thereallo.dev/blog/claude-code-prompt-steganography
The Reliability Wall -- https://arxiv.org/abs/2606.29537
Knowing When to Quit -- https://arxiv.org/abs/2606.28733
Scaling the Horizon, Not the Parameters -- https://arxiv.org/abs/2606.30616