Yesterday was the second blog post day at Refine. It came after the first week of research and iteration on incubatees ideas and questions.

It still went well. There are less posts than the first day, but even the incubatees that didn't write a post ended up working on interesting research in the process of trying to write a blog post, or in reading a lot of interesting stuff that excited them.

Here is the list of blog posts by Refine participants:

• PreDCA: vanessa kosoy's alignment protocol
• What if we approach AI safety like a technical engineering safety problem
• Benchmarking Proposals on Risk Scenarios

Here's the list of blog posts by friends of Refine and members of my Epistemology Team at Conjecture:

• No One-Size-Fit-All Epistemic Strategy
• Epistemic Artefacts of (conceptual) AI alignment research