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

Dr. Richard Moulange

πŸ‘€ Speaker
1784 total appearances

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

Because it is virology that could lead to transmissible pandemics.

594.945 View full episode β†’
80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

It's much harder for bacteria spread.

599.19 View full episode β†’
80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

You do have bacterial pandemics.

600.572 View full episode β†’
80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

I think the black death is a good example.

602.254 View full episode β†’
80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

But that was spread by vectors of fleas on rats rather than spread human to human.

604.156 View full episode β†’
80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

And so that's the virology part.

609.303 View full episode β†’
80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

The dual use part is the fact that there really are very important scientific activities that people should do in labs in careful control settings to do with making, say, influenza vaccines.

611.327 View full episode β†’
80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

So this is research that does take place every day around the world.

621.828 View full episode β†’
80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

It's dual use because they're both beneficial applications, but also relevant to misuse.

625.475 View full episode β†’
80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

And then that last bit is, it was both troubleshooting and tacit knowledge.

630.886 View full episode β†’
80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

And these are particular barriers that we at Center for Long-Term Resilience and others identified as especially important for constraining existing biological weapons attempts.

634.974 View full episode β†’
80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

Yes, I think it will be important to really get into what tacit knowledge is, the different types of tacit knowledge, how much they really are or aren't barriers to biological weapons development.

701.845 View full episode β†’
80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

But stepping back, what did VCT really do?

711.839 View full episode β†’
80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

It's a really great eval.

714.803 View full episode β†’
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 an eval that's a set of questions.

716.906 View full episode β†’
80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

And the questions are often accompanied by an image.

720.35 View full episode β†’
80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

So a lot of the benchmark is multimodal.

722.633 View full episode β†’
80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

And they'll show an image or they'll provide a paragraph that describes some sort of modern virology experiment.

726.297 View full episode β†’
80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

Maybe literally a picture of a dish with some virus in it.

732.692 View full episode β†’
80,000 Hours Podcast
AI designs genomes from scratch & outperforms virologists at lab work. What could go wrong? | Dr Richard Moulange, CLTR

And then there'll be a question like, hey, this thing looks like the wrong color or something has gone wrong with this experiment.

739.105 View full episode β†’