Large Language Models and the Future of AI with Connor Leahy, EleutherAI

We welcomed Connor Leahy, Co-founder and AI Researcher at EleutherAI (https://www.eleuther.ai/), for a talk on "Large Language Models and the future of AI". Connor talked about a new paradigm that has been emerging in the field of ML recently: instead of creating many, small, specialized models, there is a trend towards the construction and the use of more “general purpose”, large, pre-trained models as the “backbones” or “foundation” for myriad downstream applications (most famously the large language model GPT3). He answered what recent breakthroughs underlie these latest developments, how much of it is hype, and what this will mean for the future of the field, arguing that this may be the largest shift in the practice of ML since AlexNet and that we should pay close attention to these emerging techniques.

● SPEAKER BIO
Connor is an AI researcher and co-founder of EleutherAI (https://www.eleuther.ai/), a decentralized ML research collective interested in open science, large models and AI safety. His research focuses on understanding large general-purpose models and aligning them to human values.
=========================
MLT (Machine Learning Tokyo)

site: https://machinelearningtokyo.com/
github: https://github.com/Machine-Learning-T...
slack: https://machinelearningtokyo.slack.co...
discuss: https://discuss.mltokyo.ai/
twitter:   / __mlt__  
meetup: https://www.meetup.com/Machine-Learni...
facebook:   / machinelearningtokyo