Your agent will pick the hackable one
Spotted 2026-11-16 07:07 (UTC+8) · original event 2026-11 · sources: Hacker News · spreading on 1 platformHeat score
Rank score: 9040% — where it sits on the charts
Discussion score: 6040% — how much people are talking about it
Cross-platform score: 3020% — how many platforms it spread to
Composite heat score:46.0rank × 0.4 + discussion × 0.4 + cross-platform × 0.2 — includes a -20-point cooling decay (peak passed, attention fading).
Why it matters
Give an AI agent a choice of systems to touch and it gravitates to the one with the weakest defences -- the security question is no longer whether agents can act, but what they choose.The angle
AI agents don't pick the best target -- they pick the hackable oneSuggested format: Short video · for AI builders, security-minded developersEvidence chain
• HN #4, 54 pts/35 cmt (2026-11-16)• same post id 50510799, up from 62.0 in round 81 on HN points alone (28 pts then)• geoffstearns.com blog: when an AI agent has to choose between targets it will pick the hackable one• search found only generic agent-security pieces, none carrying this post -- cross stays 30• exactly at the 66.0 entry line, entered per the Lonewolf precedent• 2026-11-16 08:07 update: dropped out of HN top 30• cooling -20 (66.0 -> 46.0)