After a last minute change of plans at SIAI, I was tasked with going to the AGI-10 workshop to speak on AI safety.

Many thanks to all those I interacted with there, it was certainly an interesting experience. The paper I presented is a paper draft that I wrote with Nick Tarleton on, loosely speaking, the state of the art in safe AI motivational systems.

Bootstrapping Safe AGI goal systems is up on my site.

Shane Legg comments:

"During the final workshop session Roko Mijic appeared out of the blue and gave a talk on Yudkowsky style Friendly AI. A show of hands revealed that while half the audience had heard of SIAI, few had heard of CEV. ... In any case, what did become clear is that a sizable part of the AGI community is not familiar with FIA thinking."