voooooogel: software is defense-dominant in the AI era — attackers get leftover OSS models
voooooogel · x · 2026-10-26
In a debate with Atin, voooooogel argues software security becomes defense-dominant in the AI era: unlike bio, software can be frozen and simulated in virtual environments, letting defenders fuzz — including with agent fuzzing — millions of times before attackers ever see it.
He contrasts attacker profiles (low-income countries, cheap trailing OSS models, working around frontier-model safeguards) with defenders who have the best models, free tokens, and amortized costs — one patch protects everyone. He expects a bumpy transition but a far more secure ecosystem, likening it to fixing bugs before Y2K.
Related event: Debate: Is AI-Era Software Security Defender's or Attacker's Game?(3 posts)→
More from AGI Musings
- Why search for bipedal aliens? Maybe the universe is already full of machine civilizations — zouharvi · 2026-10-26
- Organ-Agents: 9 Organ LLM Agents Simulate 22,689 ICU Patients for Treatment Testing — mikeflache · 2026-10-26
- 56% of Organizations Cut Cybersecurity Roles Due to AI, Barracuda Report Finds — mikeflache · 2026-10-26
- McKinsey: $530B AI value pool in finance comes from replacing high-wage workers — mikeflache · 2026-10-26
- Founder proposes billboards with AI agents that call skeptics to change minds — gajesh · 2026-10-26
- Ethics book review brands moderate sentience scholar as "sacrificing humans for insects and AI" — birchlse · 2026-10-26