11 Apr 2024

Supervise the Process of AI Research — with Jungwon Byun and Andreas Stuhlmüller of Elicit

Latent Space: The AI Engineer Podcast — Practitioners talking LLMs, CodeGen, Agents, Multimodality, AI UX, GPU Infra and all things Software 3.0

This podcast episode delves into the journeys and motivations behind the founding of Ought, a non-profit research lab focused on developing AI tools for reasoning, and Elicit, a company that provides an AI research assistant. Andreas, the co-founder and CTO of Elicit, shares his fascination with AI since childhood and his desire to build practical tools that help people reason more clearly and make better decisions. The section explores the transition from research to product creation and highlights Elicit's focus on making it easier to understand existing knowledge through literature summarization and analysis. The episode also discusses the evolution of Elicit's workflow, user experience, and the evolving relationship between humans and AI agents in the context of evaluation. The importance of long-context models, list-wise re-ranking techniques, and maintaining RAG for search tasks is explored. Lastly, the episode highlights the features and use cases of Elicit, as well as the company's commitment to making research more systematic and transparent.

Takeaways

Outlines

Q & A

Preview

How to Get Rich: Every EpisodeNaval