Uploaded September 2026 | Updated November 2026, 3 days ago
What happens when AI stops merely *reading* DNA and starts learning how to *write* it? Eric Nguyen, CEO and co-founder of Radical Numerics and a key figure behind Evo, joins us to explain the leap from early genome language models to Omni—a system that can outperform specialized models on disease-variant prediction, reason across biological sequences, and even begin to “show its work” when designing better molecules. The surprising part is *where those capabilities seem to be coming from*: in several cases, the model learns useful biological structure it was never explicitly taught.
But the same capability creates a much harder question: what happens when models can design sequences that preserve biological function while looking different enough to evade today’s defenses? Eric walks us through why current biosecurity systems may be solving the wrong problem, what an AI-native defense could look like, and why the teams pushing biological design forward may also need to build the systems that defend against it. The YouTube version is especially worth watching—we put Radical Numerics’ benchmark results and model visualizations on screen as Eric walks through what they actually show.
Sponsorship and business inquiries: business@latent.space
00:00 What Is a Genome Language Model?
04:39 From Reading DNA to Writing It
07:47 The First AI-Generated Genome
11:10 Why Omni Is Different From Evo 2
24:09 Predicting Disease-Causing Mutations
39:01 Can DNA Models “Reason”?
45:18 Toward General Biological Intelligence
54:13 Designing Biology for Rare Earth Extraction
1:01:00 Why Long Context Matters for DNA
1:06:41 The Surprising Origin Story of Evo
1:12:02 Can AI Discover New Biology?
1:21:00 The Biosecurity Problem
1:25:05 Why Today’s DNA Screening May Fail
1:34:25 Is AI Biosecurity an Arms Race?
1:41:00 The Biggest Bottleneck in AI for Biology
What happens when AI stops merely *reading* DNA and starts learning how to *write* it? Eric Nguyen, CEO and co-founder of Radical Numerics and a key figure behind Evo, joins us to explain the leap from early genome language models to Omni—a system that can outperform specialized models on disease-variant prediction, reason across biological sequences, and even begin to “show its work” when designing better molecules. The surprising part is *where those capabilities seem to be coming from*: in several cases, the model learns useful biological structure it was never explicitly taught.
But the same capability creates a much harder question: what happens when models can design sequences that preserve biological function while looking different enough to evade today’s defenses? Eric walks us through why current biosecurity systems may be solving the wrong problem, what an AI-native defense could look like, and why the teams pushing biological design forward may also need to build the systems that defend against it. The YouTube version is especially worth watching—we put Radical Numerics’ benchmark results and model visualizations on screen as Eric walks through what they actually show.
Sponsorship and business inquiries: business@latent.space
00:00 What Is a Genome Language Model?
04:39 From Reading DNA to Writing It
07:47 The First AI-Generated Genome
11:10 Why Omni Is Different From Evo 2
24:09 Predicting Disease-Causing Mutations
39:01 Can DNA Models “Reason”?
45:18 Toward General Biological Intelligence
54:13 Designing Biology for Rare Earth Extraction
1:01:00 Why Long Context Matters for DNA
1:06:41 The Surprising Origin Story of Evo
1:12:02 Can AI Discover New Biology?
1:21:00 The Biosecurity Problem
1:25:05 Why Today’s DNA Screening May Fail
1:34:25 Is AI Biosecurity an Arms Race?
1:41:00 The Biggest Bottleneck in AI for Biology


![Brex’s AI Hail Mary — With CTO James Reggio (acquired for $5B by Capital One!)
From building internal AI labs to becoming CTO of Brex, James Reggio has helped lead one of the most disciplined AI transformations inside a real financial institution where compliance, auditability, and customer trust actually matter.
We sat down with Reggio to unpack Brex’s three-pillar AI strategy (corporate, operational, and product AI) [https://www.brex.com/journal/brex-ai-native-operations], how SOP-driven agents beat overengineered RL in ops, why Brex lets employees “build their own AI stack” instead of picking winners [https://www.conductorone.com/customers/brex/], and how a small, founder-heavy AI team is shipping production agents to 40,000+ companies. Reggio also goes deep on Brex’s multi-agent “network” architecture, evals for multi-turn systems, agentic coding’s second-order effects on codebase understanding, and why the future of finance software looks less like dashboards and more like executive assistants coordinating specialist agents behind the scenes.
We discuss:
* Brex’s three-pillar AI strategy: corporate AI for 10x employee workflows, operational AI for cost and compliance leverage, and product AI that lets customers justify Brex as part of _their_ AI strategy to the board
* Why SOP-driven agents beat overengineered RL in finance ops, and how breaking work into auditable, repeatable steps unlocked faster automation in KYC, underwriting, fraud, and disputes
* Building an internal AI platform early: LLM gateways, prompt/version management, evals, cost observability, and why platform work quietly became the force multiplier behind everything else
* Multi-agent “networks” vs single-agent tools: why Brex’s EA-style assistant coordinates specialist agents (policy, travel, reimbursements) through multi-turn conversations instead of one-shot tool calls
* The audit agent pattern: separating detection, judgment, and follow-up into different agents to reduce false negatives without overwhelming finance teams
* Centralized AI teams without resentment: how Brex avoided “AI envy” by tying work to business impact and letting anyone transfer in if they cared deeply enough
* Letting employees build their own AI stack: ChatGPT vs Claude vs Gemini, Cursor vs Windsurf, and why Brex refuses to pick winners in fast-moving tool races
* Measuring adoption without vanity metrics: why “% of code written by AI” is the wrong KPI and what second-order effects (slop, drift, code ownership) actually matter
* Evals in the real world: regression tests from ops QA, LLM-as-judge for multi-turn agents, and why integration-style evals break faster than you expect
* Teaching AI fluency at scale: the user → advocate → builder → native framework, ops-led training, spot bonuses, and avoiding fear-based adoption
* Re-interviewing the entire engineering org: using agentic coding interviews internally to force hands-on skill upgrades without formal performance scoring
* Headcount in the age of agents: why Brex grew the business without growing engineering, and why AI amplifies bad architecture as fast as good decisions
* The future of finance software: why dashboards fade, assistants take over, and agent-to-agent collaboration becomes the real UI
—
James Reggio
* X: https://x.com/jamesreggio
* LinkedIn: https://www.linkedin.com/in/jamesreggio/
Where to find Latent Space
* X: https://x.com/latentspacepod
* Substack: https://www.latent.space/
00:00:00 Introduction
00:01:24 From Mobile Engineer to CTO: The Founders Path
00:03:00 Quitters Welcome: Building a Founder-Friendly Culture
00:05:13 The AI Team Structure: 10-Person Startup Within Brex
00:11:55 Building the Brex Agent Platform: Multi-Agent Networks
00:13:45 Tech Stack Decisions: TypeScript, Mastra, and MCP
00:24:32 Operational AI: Automating Underwriting, KYC, and Fraud
00:16:40 The Brex Assistant: Executive Assistant for Every Employee
00:40:26 Evaluation Strategy: From Simple SOPs to Multi-Turn Evals
00:37:11 Agentic Coding Adoption: Cursor, Windsurf, and the Engineering Interview
00:58:51 AI Fluency Levels: From User to Native
01:09:14 The Audit Agent Network: Finance Team Agents in Action
01:03:33 The Future of Engineering Headcount and AI Leverage Brex’s AI Hail Mary — With CTO James Reggio (acquired for $5B by Capital One!)](https://i.ytimg.com/vi/BKLvySNVBtM/mqdefault.jpg)




![[State of MechInterp] SAEs in Production, Circuit Tracing, AI4Science, Pragmatic Interp — Goodfire
From PhD research on grounding and language models to shipping interpretability tools in production at *Goodfire,* *Jack Merullo* and *Mark Bissell* are building the infrastructure to crack open the black box—making models not just powerful, but *understandable, steerable, and safe.* We caught up with them live at *NeurIPS* to dig into the state of mechanistic interpretability heading into 2026: why interpretability is no longer just a research curiosity but a *practical tool for deployment in high-stakes industries* (healthcare, finance, life sciences), how *Goodfires platform* turns unsupervised feature discovery into real-world applications like *paint.goodfire.ai* (painting directly into Stable Diffusions mental map) and *PII detection for Rakuten* (500x cheaper than GPT-5 as a judge, higher recall than LLM-based methods), why the *memorization vs. reasoning spectrum* matters more than binary memorized or not framing (factual recall sits somewhere between rote memorization and logical reasoning), how *cross-layer transcoders and circuit tracing* are scaling interpretability across every layer of a model (creating attribution graphs through interpretable features), why *Neil Nandas pivot to pragmatic interpretability* isnt a retreat but a validation that interp is ready for real-world impact (managing by outcome, not just reverse-engineering from scratch), the *Pasteurs Quadrant* philosophy that drives Goodfire (bouncing between foundational research and applied use cases, like Pasteur pioneering germ theory _and_ vaccines), and their thesis that *interpretability unlocks latent capacity in models* that you simply cant access by treating them as black boxes—whether thats finding novel biomarkers of disease in genomics models, scrubbing PII from customer chats, or giving creative tools direct access to a diffusion models internal concept map.
We discuss:
* Jacks path: *PhD student (2020–2025) working on grounding in language models → full pivot to interpretability research → Goodfire*
* Marks path: *Palantir healthcare engineer → Goodfire in March 2024,* focused on applied interpretability and platform engineering
* What Goodfire does: *AI research company building a platform for interpreting models across modalities and domains,* focused on real-world deployment in high-stakes industries
* *paint.goodfire.ai:* using interpretability to paint directly into Stable Diffusions mental map—unsupervised concept discovery (animals, backgrounds, scenes) lets you select and paint concepts on a 2D canvas, a totally new interface for a text-only model
* Fact editing and the *ROME paper* (rank-one model editing): updating facts (capital of France is now Marseille) is an interpretability problem, not yet deployed at scale
* Whats new in 2025: *interpretability showing up in model cards, evals, red teaming exercises* (Gemini, Claude), and real production deployments like Rakutens PII detection
* AI for science: *narrowly superhuman models in genomics, medical imaging, proteomics, materials science*—interpretability unlocks novel biomarkers of disease and scientific discovery in domains humans cant natively understand (base pairs in, base pairs out)
* *Cross-layer transcoders and circuit tracing:* scaling SAEs across every layer, tying features across layers, and creating attribution graphs to trace how models produce outputs through interpretable primitives
* Why *Neel Nandas pivot to pragmatic interpretability* isnt interpretability is dead but a validation that interp is ready for real-world impact—managing by outcome, not just reverse-engineering models from scratch
* *Pasteurs Quadrant:* bouncing between foundational research (Niels Bohr understanding the atom) and applied research (Edison inventing the light bulb), with Pasteur as the model for doing both (germ theory _and_ vaccines)
—
Goodfire Team
* Goodfire: https://goodfire.ai
* Paint demo: https://paint.goodfire.ai
* Careers: https://goodfire.ai/careers
00:00:00 Introduction: GoodFires Mission and the State of Mechanistic Interpretability
00:00:56 From Grounding to Interpretability: Jacks PhD Journey
00:03:04 Paint.GoodFire.ai: Interpretability for Creative Control
00:05:30 Disentangling Memorization from Reasoning: The Spectrum of Model Capabilities
00:06:40 Unlearning vs Suppression: The Challenge of Removing Information
00:08:15 Real-World Deployments: PII Detection and Enterprise Applications
00:10:20 AI for Science: Genomics, Proteomics, and Novel Biomarker Discovery
00:11:20 Circuit Tracing and Cross-Layer Transcoders: Anthropics Contribution
00:15:44 The Neil Nanda Pivot: Pragmatic Interpretability and Managing by Outcome
00:18:37 Pasteurs Quadrant: Balancing Discovery and Application at GoodFire [State of MechInterp] SAEs in Production, Circuit Tracing, AI4Science, Pragmatic Interp — Goodfire](https://i.ytimg.com/vi/CoZp6xaTbWw/mqdefault.jpg)


