Chuck Kwong
speaker
111 appearances
1 recordings
1 series
first heard Jul 2026
last heard 20 Jul
Chuck Kwong’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 — 1 in all, peaking in Jul 2026 with 1.
Appearances
It's sort of being able to break out of your own worldview.
and assess and tap into a larger market.
That's what I love.
what we do with these evaluations.
B it's kind of like a survey in a way, where we're gathering lots of data points from a wide range of people.
And us three actually started working on consumer copilot first.
We've only recently started working on enterprise use cases.
But with consumer, there's such a breadth of people.
There's Gen Z, millennials, boomers, people with different AI influencies, different perceptions of AI.
And when you get those sort of feedback from those users, it really sort of changes.
the perception of our product teams of what you could
versus great books light because everybody has a different opinion about that.
And maybe another key comparison is also like a lot of these loss patterns are are user-driven loss patterns.
When you compare it to other types of evals or sort of like other eval resources out online, a lot of people talk about error analysis while looking at the logs or traces.
Um, so that's
kind of a very similar thing, except those error analysis is done by the product manager or done by the engineer, right?
They're looking through the logs and finding specific errors.
in the AI output, but these loss patterns are user driven.
So the users are telling us what's good or bad about the AI, but we get really specific, kind of like how error analysis is done.
But in these loss patterns, very specific things that we can fix about the model.
Showing 21–40 of 111 · page 2 of 6
← Previous
Next →