The Foretellix CTO Blog – AI safety

[Cross-posted to LW.

[Cross-posted to LW.

[LW linkpost is here] Introduction and epistemic status: This is the first post in a planned series, “Alignment as a verification problem”.

Summary: This post suggests that alignment training could benefit from coverage-driven verification.

This blog post introduces the V&V method – please read the full 26-page paper here.

From now on this blog will focus on AI safety / AI alignment (and how they connect to V&V).

My new post is here on the company’s blog.

Summary: This short post does not talk about autonomous vehicles.

Summary: This post looks further into ways to compute the residual risk of ADS.

Summary: This post talks about estimating the residual risk of an ADAS / AV design.