AI Case Studies for Natural Science Research with Bonnie Kruft

Generative AI is unlocking new research tools for bold scientific discoveries. We sort through the hype and take a deep dive into some practical examples of groundbreaking research enabled by generative AI such as small molecular inhibitors for treating infectious disease and the discovery of new materials for energy storage. As researchers reduce the discovery time from years to months, how are they ensuring that safe and responsible practices are used to instill public trust in the process?

0:00 Introduction
0:23 Scientific discovery is the most important use of AI
1:23 What Large Language Models bring to science
2:06 What makes scientific discovery different?
3:40 Prior knowledge
6:19 "No-free-lunch theorem"
9:16 Generative AI model MatterGen
9:55 Drug discovery and deep learning
14:12 Large Language Models v other training models
15:40 The evolution of generative AI models
16:37 How generative AI models can assist scientists
19:08 The role of AI in drug development
22:35 How Large Language Models can work with science-based models
26:00 Looking ahead for AI in science

[debayondharchowdhury]

The way she says "scientific"... Love it.

[BrandonMcCurry]

5:30


Exponential growth via simulated data, recurrent, repeat?

[michaelgonzalez]

Steven Meyer u are by all force of God

[rajexecutivemoreedthaniasi]

That is why, they termed as LLMs, But NOT Science, Maths Models;

[spinningaround]

No rush with Dupuytren, world needs TB cure

[Netryon]

Anything you'd wish to do depends on businesses and nobody is announcing searching for scientists in the UŽT job market. There is no reason to need anybody who already had such research, unless some dean in a faculty would know anything such sacred and won't tell about it to anybody or keep a favor for a best students stopping all the bad coming through a portal. Your value lies in things you do or work so try making websites ir your parliament just would talk about agriculture all day, but not that vast gaming you have proof of.

[MsDuketown]

I'd bet Bonnie has a hidden tattoo. Maybe she has an undiscovered dolphin somewhere?

[Corteum]

_"We believe scientific discovery is the most important use of AI"_

I strongly disagree. Scientific discovery is NOT the most important use of AI. Imo, the most important use of AI is human well being and evolution. Scientific discover should be used to suppor that, but it hasn't always been the case.

[GraczPierwszy]

Delaying alone won't do you any good.
it will only worsen your expectations,
it won't be as you want,
this is beyond your understanding,
make AI, independent AI,
uncontrolled AI,
learning from truth and real world,
and remember, once you lie, the AI will know