Latent Space: The AI Engineer Podcast From Astrophysics to AI: Building the future AI Data Stack — with Sarah Nagy of Seek.ai
18 snips
Mar 10, 2023 In this conversation, Sarah Nagy, founder of Seek.ai and former astrophysicist, dives into the fascinating world of AI and data. She explains how Seek.ai empowers non-technical users to access insights effortlessly, transforming data querying with natural language interfaces. The discussion touches on the interplay between improv and AI in crafting responses, the significance of the semantic layer for better data management, and the evolving demands in the data engineering space. Sarah also shares her predictions for rapid advancements in AI productivity tools.
AI Snips
Chapters
Transcript
Episode notes
Sarah's Diverse Background
- Sarah Nagy's background includes astrophysics, finance, and quantitative trading.
- She also pursued hobbies like classical piano, DJing, improv comedy, and theater.
LLMs and Improv Analogy
- Large language models (LLMs), like ChatGPT, can generate believable-sounding but factually incorrect information.
- This is similar to improv, where actors portray characters convincingly without necessarily being factual.
Data Backfitting Example
- A sell-side research analyst manipulated data's hyperparameters to fit a pre-existing thesis.
- This highlights the risk of misusing data to confirm biases instead of uncovering truth.