AI Leaders LeCun and Fayyad Answer Questions About the Future of AI

No items found.

June 9, 2023

Share this post

The recent Institute for Experiential AI (EAI) Distinguished Lectures seminar with legendary AI pioneer Yann LeCun, Chief AI Scientist at Meta AI (FAIR), concluded with a fireside chat and Q&A with EAI Executive Director, Usama Fayyad. In response to overwhelming interest in the seminar, we are excited to share LeCun and Fayyad's answers to many of the 150+ audience-submitted questions.

Questions in this series will include:

The trend in AI systems seems to be bigger is better. Companies keep adding more computing power and more parameters. How long can this go on, and will we asymptote? See the answer.
You’re advocating for a different approach for the AI and ML community. How long do you think it will take to pivot the field from where it is to where you would like it to be? See the answer.
What are the biggest gaps for grad students in AI education, and, in particular, the new directions AI is taking?
What are your thoughts on the recent Congressional hearings with OpenAI CEO Sam Altman? What would you have said to Congress?
If LLMs were to become perfect at language, would that give us insight into how language and natural language works? And do you think we’ll ever have an understandable LLM?
Did you actually say the “revolution will not be supervised”?
What two questions about AI do you wish you were asked more often?
Do you think quantum computing will have a significant role in the future of AI?
Given the excitement around LLMs, what are some of the understudied research directions that are possible to tackle for non-Google/Facebook-sized institutions? If you were a young investigator today, what would you do in this environment?

Watch LeCun and Fayyad discuss this topic: Why do you think the companies don’t want to talk about the importance of good, clean data and of having a human in the loop? OpenAI spent a lot of its money on curating the right balanced data sets — and the second part, which is something we’re big proponents of at the Institute for EAI, experiential AI stands for AI with a human in the loop, having that human intervention through relevance feedback.