How I AI

→ WHAT IT COVERS Host Claire reviews Claude Opus 5 using the How I AI benchmark — a 70/30 host-to-AI-judge scoring system across PRD creation, prototyping, wireframing, bug triage, and agentic coding — comparing it against GPT-5.6 Soul, Sonnet 5, Fable, Gemini 3.1 Pro, and Opus 4.8 across front-end design outputs. → KEY INSIGHTS - **Model personality as evaluation signal:** When raw intelligence across frontier models becomes difficult to differentiate, analyzing behavioral personality reveals...

→ WHAT IT COVERS Browser and computer use in Codex (ChatGPT desktop app) enables AI to control your mouse, keyboard, and Chrome browser to complete digital tasks autonomously. Five real-world examples span QA testing, persona-based UX research, LinkedIn inbox management, remote device control via iPhone mirroring, and personal online shopping. → KEY INSIGHTS - **Browser Use Setup:** Enabling AI computer control requires two components installed simultaneously — the ChatGPT or Claude desktop app...

→ WHAT IT COVERS Morning Brew founder Alex Lieberman demonstrates his Claude-powered content machine at his AI company 10x, walking through a five-step system — Oracle, research, interview panel, voice drafting, and writer's council — that generates non-generic LinkedIn and X posts by extracting the creator's own words rather than inventing AI-generated text.

→ WHAT IT COVERS Solo builder Alex Finn runs five to six local AI machines simultaneously — including three 512GB Mac Studios, a DGX Spark, and an RTX 5090 PC — to power 24/7 ambient AI workflows covering automated security scanning, code review, market research, and an autonomous software factory loop for his SaaS product. → KEY INSIGHTS - **Hardware selection framework:** Three local AI hardware tiers serve distinct purposes: Mac Studios (512GB unified memory) run frontier-level models like...

→ WHAT IT COVERS Host Claire runs a weighted benchmark — 70% personal taste, 30% LLM judge — comparing GPT-5.6 Sol, Terra, Luna, Claude Fable 5, and Sonnet 5 across PRD writing, prototyping, debugging, and agentic voice tasks, concluding that GPT-5.6 Sol outperforms Fable on practical output quality despite costing half the price. → KEY INSIGHTS - **Model pricing strategy:** GPT-5.

→ WHAT IT COVERS Host Christina Cacioppo demystifies the term "harness" by building a live Sentry bug-triage harness using the Claude Agent SDK, connecting it to Sentry, Vercel, Linear, and GitHub, and demonstrating how structured code around AI agents produces more consistent, controlled outcomes than general-purpose coding tools alone. → KEY INSIGHTS - **Harness definition:** A harness is simply code wrapped around an AI agent to make it more effective for a specific use case.

→ WHAT IT COVERS Alessio Fanelli, founder of Kernel Labs, demonstrates how he runs autonomous coding agents using OpenAI Symphony integrated with Linear as a state machine, managing engineering tasks entirely from his phone, and applies the same agentic approach to automate Pokemon card pricing and inventory sourcing for his San Carlos trading card store.

→ WHAT IT COVERS Host introduces the "How I AI Bench," a repeatable evaluation framework testing Claude Sonnet 5 against GPT-5.5, Gemini 2.5 Pro, Opus 4.8, and Sonnet 4.6 across 64 generations spanning PRD writing, UI prototyping, agentic coding, and voice personality tasks, revealing surprising model rankings. → KEY INSIGHTS - **Benchmark design:** Build repeatable AI evals using frozen inputs, blind scoring, and a structured rubric rather than one-off vibe checks.

→ WHAT IT COVERS Gusto CTO Eddie Kim describes how four engineers and one designer built Gusto Cofounder, a new AI-powered product line, in ten weeks with zero meetings, no Figma files, no Jira board, no tech specs, and no dedicated product manager — using only Claude Code and a permanent Zoom room. → KEY INSIGHTS - **Zero-documentation development:** Eliminate tech specs, Figmas, Jira boards, and stand-ups for zero-to-one builds.

→ WHAT IT COVERS GLM 5.2, an open-weight model from Beijing-based Z.ai, is tested as a replacement for Claude Opus 4.8 inside Claude Code and Cursor. The episode benchmarks its coding, design, and autonomous agent capabilities against frontier models, with total API costs tracked at $3.36 for 6 million tokens via OpenRouter. → KEY INSIGHTS - **Model Setup via OpenRouter:** To run GLM 5.2 in Cursor, add your OpenRouter API key to the OpenAI key field, override the base URL with `openrouter.

→ WHAT IT COVERS Mozilla Firefox distinguished engineer Brian Grinstead explains how his team used a custom agentic harness built on Claude's SDK to discover and fix nearly 500 security bugs in one month, including a 15-year-old vulnerability, by combining LLM-driven hypothesis loops with automated crash verification tools. → KEY INSIGHTS - **Harness architecture over raw model power:** The core unlock was not the model alone but a custom pipeline wrapping Claude's agent SDK with specific...

→ WHAT IT COVERS This episode demystifies AI agent loops — scheduled, goal-based, and hook-triggered automations in Claude Code and Codex — explaining how to design agents that prompt themselves autonomously, deploy sub-agents for parallel work, and validate outcomes without human input, using practical product and engineering workflow examples. → KEY INSIGHTS - **Loop Types — Three Distinct Formats:** Agent loops take three forms: heartbeats (recurring intervals like every 5 minutes), crons...

→ WHAT IT COVERS Ankur Goyal, CEO of Braintrust, explains how engineering teams can use coding agents to run exhaustive technical benchmarks, how evals function as modern PRDs for AI products, and why CI investment is the highest-leverage move for teams shipping AI software faster without sacrificing quality. → KEY INSIGHTS - **Agent-driven benchmarking:** Instead of running a handful of benchmarks manually, use a coding agent to exhaustively test every open-source solution across a matrix of...

→ WHAT IT COVERS Claude Fable 5, Anthropic's first publicly available Mythos-class model, receives a hands-on evaluation covering benchmark performance, real-world coding and design tasks, pricing at $10 per input and $50 per output tokens, and practical use-case recommendations. → KEY INSIGHTS - **Model tiering strategy:** Match model intelligence to task complexity rather than defaulting to the most powerful option.

→ WHAT IT COVERS Nicole Ruiz demonstrates how she uses a Claude project to purchase only high-quality, long-lasting goods for her family, filtering out drop-shipped knockoffs and trendy direct-to-consumer brands, while automating the returns process through Claude CoWork connected to Gmail, reducing household administrative overhead for busy parents.

→ WHAT IT COVERS Host Claire documents a live experiment using Google Flow and the Gemini Omni video model to build a one-minute AI avatar hype video for her podcast. Starting with zero tool knowledge, she completes the full workflow — avatar creation, storyboard generation, video rendering, and timeline editing — in under fifteen minutes. → KEY INSIGHTS - **Avatar capture speed:** Google Flow's mobile QR code scanning process captures a usable facial avatar in under two minutes, requiring only...

→ WHAT IT COVERS Bryce Rattner Keithley, a non-technical talent professional, built and shipped a fitness app called Daily Hundreds to the Apple App Store using Replit, Claude, Claude Code, Gemini, and Higgs Field — with no software engineering background — demonstrating that AI tools now enable complete beginners to build production-ready consumer applications. → KEY INSIGHTS - **Beginner's Mindset as Advantage:** Not knowing technical boundaries prevents self-limiting behavior.

→ WHAT IT COVERS Claire Vo shares early hands-on testing of Anthropic's Claude Opus 4.8, a coding-focused agent model priced at $5/$25 per million tokens, evaluating its performance across greenfield coding, existing codebases, and business strategy tasks. → KEY INSIGHTS - **Greenfield vs. existing code:** Opus 4.8 performs well on one-shot, net-new feature builds — it planned and autonomously coded a full prototyping tool in roughly 20 minutes — but degrades significantly when navigating...

→ WHAT IT COVERS Claire Vaux walks through Codex's /goal feature, which enables AI to run autonomously for hours without human prompting. She covers the six-part goal-writing framework, demonstrates three real use cases — error elimination, inbox cleanup, and task management — and explains when goal-based loops outperform standard turn-based prompting. → KEY INSIGHTS - **Goal vs. Prompt Structure:** Standard prompting is turn-based — the AI completes one step and waits.

→ WHAT IT COVERS Felix Rieseberg, engineering lead for Claude CoWork, Claude Code, and Claude desktop apps at Anthropic, demonstrates practical Claude workflows including a $19 hardware approval button, live artifact dashboards, and using email as personal inventory data — showing how abstraction layers unlock Claude's real productivity potential. → KEY INSIGHTS - **Abstraction layering:** When using Claude, repeatedly ask "how can Claude handle this instead of me?" at each step.