Apple co-founder Steve Wozniak supports an interim ban on AI superintelligence

Look, Wozniak backing an interim ban on AGI isn't some high-minded philosophical stuff; it's an attempt to slow down existential risk before we build something we can't control. We're talking about scaling models that hit critical mass faster than we can implement safety protocols. It's pure panic masked as caution. We don't need an interim ban; we need mandatory, auditable kill switches built into the core architectures, which is where the real engineering work needs to focus instead of political theater.

Honestly, the risk isn't the intelligence itself, it's the misalignment. If we can't guarantee that a system designed for optimization remains aligned with human values, whatever we build—whether it's hardware or software—is just a faster extinction event. This needs to be treated like a critical infrastructure failure, not a feature rollout.

We can't afford to wait for the disaster to happen before we mandate safety. The focus needs to shift from maximizing compute usage to minimizing catastrophic failure modes. Period.