Podcast · 103 episodes
How I AI, page 2
by Claire Vo · English
How I AI, hosted by Claire Vo, is for anyone wondering how to actually use these magical new tools to improve the quality and efficiency of their work. In each episode, guests will share a specific, practical, and impactful way they’ve learned to use AI in their work or life. Expect 30-minute episodes, live screen sharing, and tips/tricks/workflows you can copy immediately. If you want to demystify AI and learn the skills you need to thrive in this new world, this podcast is for you.
All episodes, page 2
What a harness is and how to build one with Claude Agent SDKThe episode explains “harnesses”: code that wraps an AI agent to make it more effective for a specific workflow by adding constrained context, allowed tools/actions, and required outcomes.8 Jul 2026 · 25 min · 7 chapters
How I run autonomous coding agents from my phone with OpenAI Symphony + Linear | Alessio Fanelli (Kernel Labs)How Alessio Fanelli (Kernel Labs) runs autonomous coding agents from a phone using OpenAI Symphony plus Linear as a state machine, and how he applies Codex for inventory/pricing automation in his…6 Jul 2026 · 36 min · 18 chapters
Sonnet 5 review: I ran 64 generations to find out if it's worth itThe episode reviews Anthropic’s “Claude Sonnet 5” and tests whether it’s truly “Opus-level” for cheaper pricing.30 Jun 2026 · 26 min · 9 chapters
No Figma. No Jira. No docs. How Gusto built a new product line with Claude Code | Eddie Kim (CTO)How Gusto’s CTO Eddie Kim and a tiny team built “Gusto co-founder,” a new agentic product line, in 10 weeks without traditional product artifacts (no Figma, no Jira boards, no tech specs, minimal…29 Jun 2026 · 52 min · 25 chapters
GLM 5.2: why I’m replacing Opus in Claude Code with this new modelThe episode tests GLM 5.2, an open-weight text-to-text model from Beijing startup Z.ai, as a cheaper replacement for Anthropic Opus/GPT-class models in coding workflows.24 Jun 2026 · 27 min · 8 chapters
How Claude Mythos found a 15-year-old bug in Mozilla Firefox | Brian GrinsteadMozilla Firefox’s security team explains how “Claude Mythos” (Anthropic’s model) plus a custom agentic “harness” helped find and verify long-hidden vulnerabilities, including a 15-year-old bug, and…22 Jun 2026 · 48 min · 26 chapters
How to design AI agent loops: schedules, goals, and subagents in Claude Code and CodexExplains AI “prompt loops” (automated, recurring prompting) in Claude Code and Codex: heartbeat/cron/hooks, plus “goal” loops that run until a measurable outcome is validated or blocked.17 Jun 2026 · 29 min · 12 chapters
How Braintrust uses AI agents, evals, and CI to ship better software | Ankur GoyalBraintrust’s CEO Ankur Goyal explains how coding agents, rigorous evals, and CI/observability help ship better software—especially for hard infrastructure tasks like database query latency and AI…15 Jun 2026 · 40 min · 15 chapters
Claude Fable 5 review: what the new Mythos model gets right (and very wrong)Review of Anthropic’s Claude “Fable 5” (a Mythos-class model) after early access, covering benchmark claims, safety guardrails, costs, and real-world strengths/weaknesses.9 Jun 2026 · 17 min · 7 chapters
Shopping with Claude: How to find quality brands, automate returns, and buy things that last 100 years | Nicole RuizBusy mom uses Claude (and Claude Cowork connected to email) to shop for high-quality, natural-material brands that last, automate returns, and reduce “administrative” mental load.8 Jun 2026 · 37 min · 17 chapters
Gemini Omni: Clone yourself with AI in under 15 minutesHost tests Google Flow with Gemini Omni to create an AI “clone” avatar and generate a stitched, ~1-minute hype video for the How I AI podcast in under ~15 minutes, including avatar scanning,…3 Jun 2026 · 21 min · 9 chapters
Building an iPhone app with zero technical skills | Bryce Rattner KeithleyHow Bryce Rattner-Keefley, a non-technical talent leader, built and shipped Daily Hundreds, a vibe-coded iPhone fitness app that generates customized anthropomorphic animal exercise videos using AI…1 Jun 2026 · 47 min · 18 chapters
Claude Opus 4.8 is here. Is it as good as they say?Claire Vo reviews Anthropic’s Claude Opus 4.8 coding model, asking whether it’s as good as claimed, focusing on one-shot success vs failures in edge cases, hallucinations, and strategy grounding.28 May 2026 · 14 min · 6 chapters
The Codex feature that works while you sleepHow to use Codex “Goals” (slash goal) to run long, autonomous loops that work toward an outcome, verify evidence, and stop when the goal is met—so tasks can run “overnight” without constant…27 May 2026 · 30 min · 11 chapters
How the engineer behind Claude Cowork actually uses Claude | Felix Rieseberg (Anthropic)How Anthropic engineer Felix Rieseberg uses Claude Cowork/Code to automate tedious work, keep context via connectors (especially email), and create “live artifacts” that refresh with up-to-date data;…25 May 2026 · 59 min · 26 chapters
What launched at Google I/O 2026 (30-minute day 1 recap)Day 1 recap of Google I/O 2026 launches, focusing on Gemini 3.5 models, agentic coding in Anti-Gravity, Workspace-connected Google AI Studio, consumer Gemini creative upgrades (image + video), and…20 May 2026 · 34 min · 16 chapters
HTML is the new Markdown: How Anthropic engineers are building with Claude Code | Thariq ShihiparHow Anthropic engineer Thariq Shihipar uses Claude Code with HTML instead of Markdown to create readable, interactive “rich artifacts” for specs/PRDs/plans, improving alignment and verification while…18 May 2026 · 36 min · 13 chapters
Spec-driven development: The AI engineering workflow at Notion | Ryan NystromSpec-driven AI engineering workflow at Notion—using custom agents to (1) generate high-quality standup pre-reads automatically, (2) run background coding/CI tasks from Slack, and (3) implement…11 May 2026 · 48 min · 19 chapters
Code with Claude: The 5 biggest updates explainedClaire Vaux previews five new Anthropic “Code with Claude” updates in Claude Code and Claude Managed Agents/Cloud API, plus higher usage limits.7 May 2026 · 12 min · 3 chapters
Quests, token leaderboards, and a skills marketplace: The elite AI adoption playbook | John Kim (Sendbird)John Kim (Sendbird) explains how to drive “AI-first” adoption inside a company using internal “quests” (spec-to-PRD-to-code), an AI-enabled skills marketplace (plugins/skills), and token-usage…6 May 2026 · 42 min · 20 chapters
The internal AI tool that’s transforming how Stripe designs products | Owen WilliamsOwen Williams (Stripe) explains “Protodash/Protodash Studio,” an internal AI-assisted prototyping system that generates Stripe-like, interactive product prototypes from prompts and Figma links, using…4 May 2026 · 55 min · 25 chapters
From a $6.90 newsletter to $3M API: How a non-coder built Memelord | Jason LevinJason Levin, non-coder CEO of Meme Lord, explains how he built an agent-friendly meme generator from a $6.90 newsletter into a $3M API business, and why “no UX is the best UX” for agent users.27 Apr 2026 · 52 min · 21 chapters
GPT 5.5 just did what no other model couldClara Vo reviews OpenAI’s GPT 5.5 and GPT 5.5 Pro (in Codex/ChatGPT, not yet API), claiming they’re smarter and more efficient for complex, autonomous coding.23 Apr 2026 · 24 min · 11 chapters
What Claude Design is actually good for (and why Figma isn’t dead, yet)Mini episode on Claude Design’s real-world value (especially importing design systems for consistent marketing prototypes), whether it can replace parts of the Figma workflow, and two GPT Image 2 use…22 Apr 2026 · 28 min · 9 chapters