Menu
Sign In Search Podcasts Libraries Charts People & Topics Add Podcast API Blog Pricing

Rob Wiblin

πŸ‘€ Speaker
3881 total appearances
Voice ID

Voice Profile Active

This person's voice can be automatically recognized across podcast episodes using AI voice matching.

Voice samples: 1
Confidence: Medium

Appearances Over Time

Podcast Appearances

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

You said that from your point of view, novices, amateurs, they're not where you're getting most of the uplift because I guess they're too incompetent.

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

It's overdetermined that they're going to fail even with AI assistance.

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

I couldn't help but notice in preparing for this conversation that it does seem like novices are the big focus of most of the evals that exist currently.

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

Is that because it's easier to measure because there's more novices and you don't have to pay them so much?

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

Yeah, so I would have thought you would expect that the more sophisticated actors would get a bigger... Just to explain, so the mental model that I've had as a result of thinking about this for one minute is you've got kind of an S-curve with all of these things.

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

I'm gesturing up here, kind of making an S. I guess on the X-axis, you've got...

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

Just how much expertise do you have in the area?

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

And then what's your probability of success?

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

So there's a point at which you can already do it, in which case you don't need the AI to help you.

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

There's a point at which you're doomed to failure, no matter how much someone coaches you, because you're just no good.

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

And it's the people in the middle who you would think would get the biggest boost from having some advice.

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

You are exactly right.

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

I mean, I think something that's slightly useful about thinking about the S-curve is that the S-curve is going to differ depending on the challengingness of the thing that the person or the group is trying to do.

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

So if we're thinking about making mirror bacteria, something that no one has ever done, and that is actually...

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

some of the most challenging frontier science possible, then probably the only group that would be meaningfully helped would be a state actor like the Russian bioweapons program.

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

They're the only folks who would be close enough to having a shot that AI assistance would help them out.

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

I guess I was excluding them because I'm imagining that they wouldn't do it.

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

Correct.

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

Because they're not typically bioweapons programs.

80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

And I guess conversely for like the absolutely most basic, I suppose perhaps like chemical weapons attacks that are more straightforward than biological weapons attacks, it might be the novices who now are getting the biggest uplift because they were the ones who would struggle.