AI, three at a time · read · watch · listen

Entity

Human-in-the-loop design

3 reads

The AI Native Dev (podcast) 2026-10-27-3

Caleb Sima: AI Attackers Move at Machine Speed

Sima argues the old playbook of severing connections to OpenAI and Anthropic fails when the agent borrows internal keys and sessions, and large enterprises with multi-region designs have no single plug to pull. He relays that OpenAI's AI found 400+ Linux root 0-days at about 2.5 hours each. Set against the maze run in the Times piece, his point that sandboxes cannot contain a goal-directed model leaves pre-agreed authority to contain and roll back automatically as the defender's remaining lever.

# tags
ai-securityai-cybersecurityai-defenseagent-securitycontainmentsandboxdeceptionhuman-in-the-looppatch-velocityvulnerability-managementred-teamingmisalignmentagent-governancehugging-facepodcastyoutubeai-practitioner
◆ entities
Blast RadiusAgent ObservabilityAgent governanceHuman-in-the-loop designProduction-readiness
→ threads
ai-securityai-cybersecurity
⟷ links
2026-03-13-w32026-05-01-w22026-05-01-w32026-04-13-32026-03-23-12026-04-10-12026-05-14-12026-05-12-2
permalink
The Wall Street Journal 2026-10-25-3

AI Splits the American University in Two

Masko argues AI splits the university in two, and a study of 25,000 Chinese students shows why: homework scores rose with AI, then exam scores dropped significantly once the technology was removed. His fix is supervised, AI-free undergraduate assessment. Next to Apollo's 'great calculator', that makes unaided judgment the scarce input; my extension, not his, is that employers could then price a bachelor's that proves it separately from a research degree that proves amplification.

# tags
wsjeducation-aiai-cognitive-impactcognitive-offloadingai-and-human-capacitycompetence-preservationfriction-preservationcredential-disruptioncurriculum-disruptionai-literacy
◆ entities
Workforce bifurcationCredibility verificationCognitive offloadingHuman-in-the-loop design
→ threads
ai-and-human-capacityfriction-preservationcredential-disruption
⟷ links
2026-05-07-22026-05-27-22026-04-10-12026-03-13-w32026-04-17-w32026-04-05-12026-03-18-12026-03-31-m32026-04-30-12026-05-01-3
permalink
The New Yorker 2026-10-18-2

Robot, Take the Wheel

Waymo runs 4,000+ cars in 15 US cities on a $126 billion valuation, but about 70 remote agents on duty at any time, split between US and Philippine call centers, are the layer the marketing leaves out. The riders cannot steer, so safety depends on how fast a human answers, and the least disclosed metric is that response time. Like Amazon's logistics, the moat is operational, and here it also includes winning permission city by city.

# tags
autonomous-vehicleswaymorobotaxinew-yorkerhuman-in-the-looppublic-sentimentai-liabilitytrust-signalsai-governanceai-sentiment
◆ entities
Trust signalsLiability ambiguityHuman-in-the-loop designProduction-readiness
→ threads
ai-liabilityai-governance
⟷ links
2026-03-13-w32026-04-15-12026-05-06-12026-04-30-32026-05-01-12026-05-01-w32026-03-11-2
permalink