Yoshua Bengio
speaker
2,216 appearances
5 recordings
5 series
first heard Oct 2018
last heard 7 May
Yoshua Bengio’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 — 3 in all, peaking in May 2026 with 1.
Appearances
to use this buffer about where does the system make errors that are too large?
The mathematical guarantees arise from a different source.
They come from... So first of all, the form of the mathematical guarantees is that either the predictor or the agentic version will...
have a exponentially small probability of achieving what I call a challenging and harmful goal.
So what do I mean by this?
Anything that a randomly initialized neural net would not be able to do, except if you're like incredibly unlucky, is something you're protected against, right?
So it's a very strong protection.
What evil can come from a randomly initialized neural net?
Not much, right?
This is the level of guarantee.
Now, it's not 100%.
It's a lot better than what we have now.
But it's like, you know, two to many thousands.
And, you know, it's very, very unlikely to the point where it's like astronomically unlikely.
But that's the kind of guarantee you get.
And
The reason you're getting those guarantees is because, well, first you start with an initialized network, an initialized network that is incapable, but the training objective then pushes away from bad behavior.
And the reason it pushes away from bad behavior is that there's this guardrail system.
in order for an AI to be able to achieve something bad, it's going to have to deviate quite a bit from the Bayesian predictor, which is a target of training.
And those deviations are penalized by the training objective.
Showing 181–200 of 2,216 · page 10 of 111
← Previous
Next →