Cal Newport
speaker
22,298 appearances
39 recordings
5 series
first heard Oct 2025
last heard 10 Sep
Cal Newport’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 — 39 in all, peaking in Jun 2026 with 9.
Appearances
So here's my policy proposal, which is not really a real one because I don't know how to actually put this, you know, the regulatory language.
podcast-host
But let me just give you like a blue sky thought experiment.
podcast-host
Imagine we basically said you can't have long horizon LLM driven actuated agents.
podcast-host
Programs whose action is largely determined by the result of LLM prompts without human supervision.
podcast-host
And the way you would do this is you would have a depth limit.
podcast-host
That type of system can do like half a dozen prompts unsupervised and that's it.
podcast-host
I don't know if that's the right number, but something like that.
podcast-host
So I could do a few prompts to try to solve a very immediate thing you asked it to do, but none of this long horizon stuff.
podcast-host
Just make that illegal.
podcast-host
You know what?
podcast-host
I think the AI companies would welcome it because the only reason, I don't know if this is true, another hot take, but I think the only reason why they're running these experiments with these long horizon LN prompt loops is because it's marketing, basically.
podcast-host
right?
podcast-host
Like it's, they have these benchmarks for like a meter made this, the AI evaluation nonprofit made this challenge benchmark about these challenges that required many, many steps.
podcast-host
And they use, they plot each LLM on this scale.
podcast-host
Well, you can't, an LLM alone can't do anything, but it's LLM plus some sort of harness that they give and they let it run for a long time.
podcast-host
The challenge is how far can you go unsupervised?
podcast-host
In cybersecurity, the exploit gym challenge, this benchmark is about
podcast-host
How much can you hack unsupervised without any humans?
podcast-host
And so the company's like, we got to do good on these benchmarks because it feels sci-fi.
podcast-host
It feels high tech.
podcast-host
Showing 361–380 of 22,298 · page 19 of 1115
← Previous
Next →