training data

3 statements across 3 episodes · 0 bullish · 1 bearish · 3 people on the record · first statement Nov 25, 2024 by Lin Qiao · across every show →

Everything said about training data, oldest first

Nov 25, 2024 neutral
Insight
Customer inference workloads rarely align with foundation model training distributions
“The data distribution in their inference workload doesn't align with the data distribution in the training data for the model, right? It's a given, actually. If you think about this, because researchers have to guesstimate what is important, what's not importa…”
Lin Qiao Nov 25, 2024 ▶ 15:29 Why Compound AI + Open Source will beat Closed AI — with Lin Qiao, CEO of Fireworks AI
Jan 10, 2025 neutral
Insight
Bryk: Search engines do not need PhD-level training data
“With search, you're asking, like, simple questions about billions of things, like, is this a startup? Did this person write a blog post about search? You know, those are actually simple questions. You don't need, like, PhD level training data.”
Will Bryk Jan 10, 2025 ▶ 49:13 Beating Google at Search with Neural PageRank and $5M of H200s — with Will Bryk of Exa.ai
Feb 12, 2026 negative
Assertion Not checkable as stated
Most Protein Design Validation Uses Targets With Training Data Overlap
“One of the things that, you know, we found, ah, with the field was that a lot of the validation, especially outside of the validation that was done on specific problems, was done on targets that have a lot of, you know, known interactions in, in the training d…”
Gabriele Corso Feb 12, 2026 ▶ 1:19:49 🔬Generating Molecules, Not Just Models
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 200 episodes transcribed, diarized, checked and made playable, runs on the StarZero media pipeline. Drop in your own episode and the podcast clipper finds the moments worth sharing, cuts them, captions them, and reframes them for every feed.