Dr. Richard Moulange

speaker
1,784 appearances 1 recordings 1 series first heard Mar 2026 last heard 31 Mar

Dr. Richard Moulange’s voice in public audio — every appearance, attributed to the second.

Trend

recordings per month · last 12 months
1 · Mar OctJan 26AprJulnow

Recordings per month over the last 12 months — 1 in all, peaking in Mar 2026 with 1.

Appearances

newest first · ▶ plays the moment
80,000 Hours Podcast · #241 – Richard Moulange on how now AI codes viable genomes from scratch and outperforms virologists at lab work — what could go wrong? · 31 Mar 2026 podcast
And so this is why I think a lot about autonomy and tool use and who is actually going to be doing the step.
Is it the human or is it the AI?
Because the more steps the AI can do, the larger the number of actors that can in fact attempt things that otherwise they would be bottlenecked.
Oh, I don't want to talk about Claude too much, but I think it's also relevant.
The cyber attack that was instigated, Anthropic allegedly was instigated by Chinese-sponsored cyber attackers that use Claude code.
It was noted that something like 90% of the attack was done autonomously because the humans didn't have to intervene too much.
That's not just a time save, it's actually an enabler.
Because if you had to keep jumping in and to fix it at every moment, you wouldn't do it in the first place.
And so that's why I am especially concerned by autonomy, by automatic tool use, because I think it will turn on risk chains, as they call it.
It will make possible the sorts of sequence of activities that previously threat actors just wouldn't bother trying.
I have sometimes been shocked by meeting colleagues in AI safety
who just sort of dismiss bio as a totally irrelevant domain.
I think this comes in two flavors.
One is to say it's just irrelevant compared to something like misalignment.
It's just not important.
Misalignment is what will happen about whether AI takes over the galaxy or not.
Who cares about biological weapons?
That is sort of a very human concern.
I think this often comes from people thinking that AI bio risks are only human-led.
But in fact, we should be thinking about AIs themselves, developing and deploying biological weapons.
Showing 461–480 of 1,784 · page 24 of 90 ← Previous Next →