/
© 2026 RiffOn. All rights reserved.

Get your free personalized podcast brief

We scan new podcasts and send you the top 5 insights daily.

  1. Latent Space: The AI Engineer Podcast
  2. 🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science
🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science

Latent Space: The AI Engineer Podcast · Sep 22, 2026

Google's John Platt on ERA, an AI that evolves code to solve science problems, and the shifting role of scientists in an age of automation.

Agentic AI Succeeds Where Evolutionary Coding Failed By Using Intelligent, Not Random, Code Mutation

Past attempts at evolutionary programming failed because random code mutations are mostly useless. Modern agentic systems like Google's ERA work because the "mutation" is guided by an intelligent LLM with vast world knowledge, which proposes sane, high-potential changes.

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science thumbnail

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science

Latent Space: The AI Engineer Podcast·2 months ago

Automated Science Systems Are Fundamentally Constrained by Goodhart's Law

Systems like Google's ERA, born from the idea of automating Kaggle competitions, constantly run into Goodhart's Law. When a metric becomes the optimization target, it's easily gamed and ceases to be a reliable measure, requiring constant human oversight and re-evaluation.

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science thumbnail

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science

Latent Space: The AI Engineer Podcast·2 months ago

AI for Science Must Aim for Descriptive, Not Just Predictive, Models to Enable Extrapolation

A predictive model minimizes error on a dataset (e.g., "apples fall") but can't extrapolate. True science requires descriptive models that capture underlying physics (e.g., gravity), allowing for extrapolation to new domains like the movement of planets.

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science thumbnail

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science

Latent Space: The AI Engineer Podcast·2 months ago

Powerful AI Tools Demand More, Not Less, Scientific Rigor to Avoid Overfitting

Agentic AI can explore a vast hypothesis space, massively increasing the risk of overfitting. This requires scientists to be more rigorous than ever, using techniques like hidden holdout datasets to avoid being fooled by the tool's power, which can "slice your fingers off."

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science thumbnail

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science

Latent Space: The AI Engineer Podcast·2 months ago

AI Revolutionized Weather Forecasting But Stumbles on Climate Modeling Due to Data Scarcity

AI has transformed short-term weather forecasting (a data-rich, interpolative problem). However, it has not yet revolutionized long-term climate modeling, which is a data-poor, non-stationary problem requiring extrapolation where we inherently lack future data.

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science thumbnail

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science

Latent Space: The AI Engineer Podcast·2 months ago

Google's ERA Reframes Scientific Discovery as Defining 'Scorable Tasks' for AI Agents

Google's AI for Science tool, ERA, maps complex scientific problems into tasks where an AI agent generates code to maximize a specific score. This shifts the scientist's role from coding to defining an objective function, representing a meta-level change in the research process.

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science thumbnail

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science

Latent Space: The AI Engineer Podcast·2 months ago

AI Forces Scientists to Specialize in Either 'Creativity' or 'Rigor'

As AI automates coding and analysis, the role of a researcher is bifurcating. Some will focus on the 'creative source'—generating novel hypotheses and scientific directions. Others will become the 'rigor source,' ensuring the AI's output is correct, scalable, and trustworthy.

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science thumbnail

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science

Latent Space: The AI Engineer Podcast·2 months ago

Modern Transformers Are a Sophisticated Realization of 1980s Hopfield Networks

John Hopfield's work on associative memories (Hopfield networks) in the 1980s was a foundational concept for neural networks. John Platt highlights that today's dominant Transformer architecture is, at its core, a highly advanced form of that same associative memory concept.

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science thumbnail

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science

Latent Space: The AI Engineer Podcast·2 months ago

The Final Frontier for AI in Science is an Automated 'Everything Lab'

The ultimate bottleneck for fully automated scientific discovery is performing physical experiments. The ideal breakthrough would be a universal "everything lab" that could execute any physical experiment on demand, closing the loop between computational hypothesis and real-world validation.

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science thumbnail

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science

Latent Space: The AI Engineer Podcast·2 months ago

Complex AI Tools Cross a "Phase Transition" in Efficacy With New LLM Versions

The effectiveness of sophisticated AI systems doesn't scale linearly with LLM improvements. Instead, they hit a capability threshold. John Platt notes his ERA system was "impossible" with Gemini 2.0 but became "amazing" with version 2.5, demonstrating a phase change in utility.

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science thumbnail

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science

Latent Space: The AI Engineer Podcast·2 months ago

Google's AI Agent Solved a 2-Year Climate Research Bottleneck on Contrails

Researchers were stuck for two years on how to measure the warming effect of jet contrails, a difficult counterfactual problem. The ERA system successfully searched through potential confounders to generate a working model, unblocking a key scientific challenge.

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science thumbnail

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science

Latent Space: The AI Engineer Podcast·2 months ago

AI's Automation of Scientific Grunt Work Threatens the Development of Researcher 'Taste'

A major pedagogical challenge posed by AI in science is how the next generation will develop deep intuition. If AI handles foundational tasks, it's unclear how young scientists will build the "taste" that traditionally came from personally struggling with those very problems.

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science thumbnail

🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science

Latent Space: The AI Engineer Podcast·2 months ago