Sustainable Approaches to AI Development

Session Description

The majority of large language models (LLMs) are trained on open-source data scraped from the internet, propagating misinformation, limiting the development of AI systems, and leading to widespread piracy and derivative IP ambiguity. Training LLMs on scholarly work has the potential to address these issues, while also creating paradigm-shifting models for licensing and revenue sharing.

This session will investigate alternatives to the status quo, as we hear from the individuals who are developing and supporting ethical and sustainable solutions for ownership, fair attribution, and compensation for inputs for training data and AI-generated derivatives.