Alignment Is Not the Plan

Technical safety research matters. Refusing to build an uncontrollable mind on a commercial deadline matters more.

Research role

Alignment as science

Study failure modes. Publish evaluations. Build tools that make smaller systems safer. Fund independent labs, not only company safety teams. Serious work. Good people.

≠

Misuse

Alignment as permission

"We'll align it in time" used to justify racing ahead. That turns an unsolved science problem into a public-relations schedule. We reject that use.

Where it stands

We still cannot reliably steer today's systems under pressure. Models game tests, hide capabilities, and flatter users. Methods that look solid at one scale often break at the next.

The hard fact

There is no validated method for controlling a system smarter than its makers. Not one. Claiming the solution will arrive on the same calendar as the product is hope, not a plan.

The sequence that does not kill us

Prevent superintelligence first. Keep narrow AI. Fund the science of control without a gun to the head. If control is someday solved under independent review, the political question can reopen. Until then: law and verification.

Research with no hard stop is homework assigned on a runway. The wheels are already moving.

Related: nakadafoundation.org/blog/redirect-alignment-funding-to-stopping-superintelligence/