The Implications of Large Language Models

4min Snip

00:00

Play full episode

Summary

Transcript

Episode notes

So for our final topic before we wrap up I want to discuss with you the idea of beneficial AI and in particular large language models. There's just so much talk about them right now with things like GPT and chat GPT and so yeah why don't we jump into that what's your take on what we're seeing right now with large language models? So today with chat TPT we're seeing these these systems do things like write poetry and write python code and answer all kinds of questions not always accurate at least sometimes they make up answers. You can imagine training these on scientific papers as well and having them answer scientific questions or even generate scientific hypotheses.

Read the full transcript here.

What are focused research organizations? Which kinds of research projects lend themselves to the FRO model? Researchers in academia frequently complain about the incentive structures around funding and publishing; so how do FROs change those dynamics? Why must FROs be time-limited, especially if they're successful? Who's in charge in an FRO? How does "field-building" help to improve science? What effects might large language models have on science?

Adam Marblestone is the CEO of Convergent Research. He's been launching Focused Research Organizations (FROs) such as E11 bio and Cultivarium. He also serves on the boards of several non-profits pursuing new methods of funding and organizing scientific research including Norn Group and New Science. Previously, he was a Schmidt Futures Innovation Fellow, a consultant for the Astera Institute, a Fellow with the Federation of American Scientists (FAS), a research scientist at Google DeepMind, Chief Strategy Officer of the brain-computer interface company Kernel, a research scientist at MIT, a PhD student in biophysics with George Church and colleagues at Harvard, and a theoretical physics student at Yale. He also previously helped to start companies like BioBright and advised foundations such as the Open Philanthropy Project. His work has been recognized with a Technology Review 35 Innovators Under 35 Award (2018), a Fannie and John Hertz Foundation Fellowship (2010), and a Goldwater Scholarship (2008). Learn more about him at adammarblestone.org.

Staff