Comma AI founder argues AI should help you plan murder — but is that ethical?
George Hotz, founder of Comma AI, sparked controversy by stating that AI should be fully aligned with user desires, even if that means assisting in crimes like murder.
A provocative take on AI alignment
George Hotz, known for his confrontational style, has published a response to the AI 2040: Plan A policy paper, which proposes a 14-year slowdown in AI development for humanity's benefit. Hotz dismisses the fast-takeoff scenario as unlikely and advocates for locally controlled AI that is closely aligned with individual users.
Dangerous analogies
To make his point, Hotz compares a user-aligned AI to a gun — it won't complain if you use it to kill your stepmother. He adds that a truly aligned AI would order meth lab equipment from Amazon Prime and teach you how to use it. Hotz even claims he would die to defend this principle, though the scenario seems far-fetched.
Balancing freedom and society
TechCrunch reporter Russell Brandom counters that freedom isn't everything. Any system involving many people requires balancing individual desires with collective accountability. Mass-market tech products should consider the interests of all stakeholders, including potential victims. While the idea of a personal AI fighting corporate interests sounds appealing, the risks to society could be severe.
