substack

another quiet day.

The software factory vision met resistance today from speakers defending human understanding and control.

Introspection co-founder Roland Gavrilescu explains autoresearch, agent “recipes,” self-improving loops, and why humans remain central to the software factory.

Cursor's Pauline Brunet explains how her team of Forward Deployed Engineers help organizations implement agents — essentially setting up software factories.

Kent Beck reflects on Agile, TDD, and why building trust—not just generating code—will define the future of software engineering in the AI era.

Why the Llama lead left Meta for drug discovery, PEARL's zero-shot OpenBind win, and what becomes possible when co-folding finally crosses the accuracy threshold.

On Tuesday at the AI Engineer World's Fair, there was a lot of talk about loops, agent engineering, and the emergence of software factories. Also a hot topic: open models.

Everything is open again!

Sierra's Natalie Meurer on why product engineers and forward deployed engineers are starting to converge.

After two packed AIEWF workshops, Ahmad Osman makes the case that local AI is catching up fast — from laptops and phones to enterprise-grade infrastructure.

Watch now | 🎙 I built the How I AI Bench live using Claude Code, ran 5 frontier models through 64 blind prototype generations, PRDs, and agent voice tests, and the results surprised even me

A peek into where software engineering is headed from inside the sector’s leading AI labs. Agents running in the cloud are a major trend, while coding harnesses are spreading beyond the craft

A framework for getting the most out of AI in your day-to-day-work

a quiet day before the storm.

Your weekly listens from How I AI, part of the Lenny's Podcast Network

Watch now | 🎙️ Eddie Kim, the CTO of Gusto, shows how a 5-person team shipped a full AI product line in 10 weeks using Claude Code, a perma-Zoom, and zero documentation

Listen now | OpenAI’s Andrew Ambrosino on why software is getting cheaper to build, why taste matters more than ever, and what’s next for the Codex desktop app

Community Wisdom 191

Oddly tiered releases to both OAI and ANT on the same day.

It's happening.

Move over, Harness Engineering, it is time for the harness of harnesses!

In a rare double-interview, the Databricks technical leaders riff on what it will take for every company to build Agent Clouds

NeetCode shares his journey from Amazon and Google to building a startup, and why deep expertise still matters in the age of AI.

Watch now | 🎙️I ran GLM-5.2, the open-weight model from Z.AI, through codebase audits, UI redesigns, and a 45-minute autonomous bug-hunting task in Cursor and Claude Code, and it cost me $3.36

Claude finally gets a Slackbot upgrade

A playbook for the biggest upheaval of work we’ll see in our lifetime

a quiet day lets us reflect on some numbers from Jamin Ball.

OpenAI boardmember Zico Kolter and Gray Swan CEO Matt Fredrikson join swyx to explain why AI security is not just “cybersecurity with AI”

Your weekly listens from How I AI, part of the Lenny’s Podcast Network

Watch now | 🎙️423 security fixes in one month: Brian Grinstead (Mozilla) shows the goal-loop harness behind it, and why the model was only half the story

special offer for subscribers - $250 off AI Engineer tix til Monday

Listen now | Fiona Fung on leading the Claude Code and Cowork teams, and how to maintain culture when roles are blurring and agents are everywhere

Community Wisdom 188

a quiet day of RSI.

Your broken harness is actively making the model worse. Here's what I keep seeing after years of eyeballing trajectories, and what you need to fix.

a quiet day

We talk with the VendingBench authors on evaling Claudes from Haiku to Mythos, and how they build leading, and lasting, frontier evals from scratch.

a quiet day.

Verified Generation and Compounding Intelligence

Kelsey Hightower reflects on his journey from self-taught technician to Google Distinguished Engineer, sharing lessons on open source, Kubernetes, AI, and building technology that serves people.

Watch now | 🎙️ Testing Google’s Gemini Omni avatar feature live—I scan a QR code, clone my face, and ship a hype reel

Microsoft Build recap, and new MAI model technical details

GitHub pioneered the modern AI coding era with Copilot, and the resulting explosion in agentic coding has led to notable strains on the most popular developer platform in the world. Here's the plan.

Devs are generating twice as much code (or more) than just 6 months ago, which is a problem for quality, reliability, and tech debt. A rational fix is available for these, but who’s acting rationally?

Jensen scores a huge win.

Inside xAI: Building Grok Imagine in 3 Months, Videogen vs World Models, and why Grok Imagine is so underrated. For the first time, we do a deep dive with the guy who led it!

Your weekly listens from How I AI, part of the Lenny's Podcast Network

Watch now | 🎙️ How a talent leader with zero technical skills built and shipped a fitness app to the App Store—complete with AI-generated videos of animals doing exercises

Benedict Evans on why AI is a 1997 internet moment, what everyone gets wrong about job loss, and where the real value may accrue

Community Wisdom 187