Cities (261) ▾

AI Tinkerers - "One-Shot"

How to Ship Complex Features 10x Faster with AI Agents | Dex Horthy (HumanLayer)

How to Ship Complex Features 10x Faster with AI Agents | Dex Horthy (HumanLayer)

March 04, 2026 · 2 minutes

In this episode, Dexter Horthy (HumanLayer) explains the 12 Factor Agents framework for building multi-step agentic workflows with structured outputs and simple primitives—bypassing heavy orchestration. He shares production lessons from nine months of running Dex’s Claude Code, including Riptide’s step-based architecture, Linear ticket integration, GitHub artifact storage, and autonomy levels from human review to serverless GitHub Actions.

Read post →
What Happens When You Hit Claude’s Limits? | Sam Hesson (Meta AI)

What Happens When You Hit Claude’s Limits? | Sam Hesson (Meta AI)

February 25, 2026 · 3 minutes

Sam Hesson (ex-CTO/co-founder of Nanome) explains what “maxing out” Claude’s limits looks like: $50,000 in token spend over 2.5 months that led Anthropic to rate-limit Claude Max. He built an automated CI/CD “Darwinian” system: three parallel agents on separate VMs submit competing PRs, then an LLM judge with a rubric selects the winner.

Read post →
The Architecture of Vibe Coding: Inside Bolt.new's Stack

The Architecture of Vibe Coding: Inside Bolt.new's Stack

January 26, 2026 · 3 minutes

In a San Francisco chat, Eric Simons (StackBlitz) explains how Bolt enables “vibe coding” with WebAssembly/WebContainers: a browser-native, near-instant Node.js-like environment that avoided cloud-VM latency, fueled ~7 years of IDE work to ~$700K ARR, and is shifting engineering from writing specs/code to designing systems, reviewing agent output, and hardening security/enterprise workflows.

Read post →
This is a promotional graphic, likely a video thumbnail, featuring a headshot of Lukas Biewald against a dark, circuit-themed background, advertising an interview about tinkering, observability, and building AI companies.. Text: LUKAS BIEWALD ON TINKERING, OBSERVABILITY, AND BUILDING TWO AI COMPANIES AI TINKERERS TECHNICAL INTERVIEW AI TINKERERS “ONE-SHOT”

Lukas Biewald on Tinkering, Observability, and Building Two AI Companies

December 18, 2025 · 3 minutes · 2 photos

In this AI Tinkerers One Shot, Lukas Biewald—founder of Weights & Biases and CrowdFlower (now Figure Eight)—explains how early tinkering (robot cars, Raspberry Pi experiments, meeting Pete Warden) shaped a culture of experimentation that fuels AI tooling. Key takeaways: agentic coding workflows, production observability, and where fine-tuning/RL are headed.

Read post →
Steve Yegge on Agentic Coding, Beads, and the Future of AI Workflows

Steve Yegge on Agentic Coding, Beads, and the Future of AI Workflows

December 01, 2025 · 1 minute

Steve Yegge, ex-Sourcegraph (Amp) and longtime Amazon/Google engineer, shares real agentic-coding workflows: the 85 percent rule, how to guide agents like a team lead, the “Land the Plane” session shutdown protocol, Beads (a Git-backed memory layer for structured work), and why multimodal debugging is becoming essential—watch now: https://youtu.be/s96O9oWI_tI

Read post →
Building Robots That Feel Alive: Inside Innate’s Mars Launch

Building Robots That Feel Alive: Inside Innate’s Mars Launch

September 07, 2025 · 2 minutes

Innate’s MARS is a 25×25×40 cm open-source tabletop robot that plays chess, learns via a community, and serves as a social lifeline; it runs NVIDIA Jetson, uses Wi‑Fi/WebSocket for real-time control, and a Python SDK on open hardware with modular sensors. Three frontiers: manipulation, agentic real-world frameworks, open-source democratization.

Read post →
Rebuilding GPT-2 in a Spreadsheet — Ishaan Anand on Opening the Black Box

Rebuilding GPT-2 in a Spreadsheet — Ishaan Anand on Opening the Black Box

August 14, 2025 · 3 minutes

Ishaan Anand rebuilds GPT-2 from scratch inside a spreadsheet, exposing every weight, bias, and operation to visualize tokenization, embedding, transformer blocks and attention in real time; ~600 lines of Python replicate GPT-2 behavior, with adjustable temperature, top-k, top-p. Also highlights innovations since GPT-2: Mixture of Experts, RoPE, training advances.

Read post →
Events Gallery FAQ Sponsor