Zershaaneh Qureshi
speaker
853 appearances
6 recordings
1 series
first heard May 2026
last heard 17 Sep
Zershaaneh Qureshi’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsRecordings per month over the last 12 months — 6 in all, peaking in Aug 2026 with 2.
Appearances
Um, there has been
Some recent anthropic research on teaching Claude Y, which does shed a little bit of light on this, but fuels kind of early stage.
What I'm interested in here is how much symmetry you expect there to be between.
this phenomenon and your emergent misalignment results.
Yeah, yeah.
So there's there's one way to be perfect, but there are lots of ways to be bad.
Um, so there's sort of fundamental asymmetry there.
And then yeah, you say this thing about kind of examples, um
in its data where it's kind of like, yeah, there are lots of bad archetypes, but there's not really an archetype of like like the perfect AI model that it kind of exists anywhere that we can point it to, right?
Yeah, yeah.
So I'm I'm getting the sense that you don't have
High hopes for
a research project that tries to kind of leverage emergent misalignment to align our AI systems, right?
I'm kind of wondering
what we'd need to understand better or become more confident about, um, to be able to kind of like use this as as more of a lever in our um in training future models.
So despite all of these results that we've been talking about, um, which have mostly been in experimental conditions, I think somebody might say, look, overall it seems like we're doing a pretty good job of aligning AI already.
You know, chatbots like the latest iterations of Claude and Chat GPT are broadly doing a pretty good job of
Doing what we want them to do, not doing terrible things, um, and sort of seeming broadly aligned um on the surface.
Is that the impression that your research gives you?
Because I guess if so, then maybe there's an argument that we should kind of just keep doing what we're doing and and hang tight.
Showing 421–440 of 853 · page 22 of 43
← Previous
Next →