Sander Schulhoff
speaker
947 appearances
1 recordings
1 series
first heard Jun 2025
last heard Jun 2025
Sander Schulhoff’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsNo recordings in the last 12 months.Older appearances are listed below; set an alert to hear about the next one.
Appearances
So I I don't know how realistic those are.
Well, you can train the model on those laws, but You can still trick it.
You can still trick it.
So there there is hope, but we have to be kind of realistic about where that hope is and who is solving the problem.
Uh and it has to be the AI research labs.
Uh, you know, there's there's no like
like external product focused companies really, oh, you know, I have the best guardrail now.
It's not a realistic solution.
It has to be the AI labs.
Uh it has to be, I think it has to be innovations in model architectures.
Uh I've seen some people say like
Oh, you know, like
Humans can be tricked too, but I feel like the reason we're so sorry, these these are not my words to be clear.
Um, the reason that we're so uh able to detect like scammers and and other uh bad things like that is that we have consciousness uh and we have a sense of self and not self.
And it could be like, oh like
Am I acting like myself or like this is not a good idea this other person gave to me?
Uh and kind of reflect on that.
Uh I guess, you know, LMs can also kind of self-criticize, self-reflect.
But I've seen consciousness proposed as a solution to prompt injection, jailbreaking.
Not like a hundred percent on board with that, not entirely on board with that, but I I think it's interesting to think about.
Showing 741–760 of 947 · page 38 of 48
← Previous
Next →