FARI Research hosts the panel on model dependability gaps
Confirmed for Seoul Alignment Workshop: @steverab (@Princeton), on the opening panel on the capability-reliability gap. He works on when machine learning systems can be trusted: uncertainty, calibration, knowing when a model should abstain. The premise underneath it: a capable system and a dependable one are not the same, and the gap has to be measured, not assumed.
Users in the replies express enthusiasm for the Seoul AI Alignment Workshop addressing the capability-reliability gap because they work in the same topic from a cognitive ergonomics perspective and want to connect.
No Digg Deeper questions have been answered for this story yet.
Very excited to discuss the capability-reliability gap in agentic AI at the @farairesearch Seoul Alignment Workshop next week!
Confirmed for Seoul Alignment Workshop: @steverab (@Princeton), on the opening panel on the capability-reliability gap. He works on when machine learning systems can be trusted: uncertainty, calibration, knowing when a model should abstain. The premise underneath it: a capable system and a dependable one are not the same, and the gap has to be measured, not assumed.
@steverab @farairesearch Nice ! I work in the same topic but from a cognitive ergonomist’s pov. Let’s connect ?
