Sources
- AI Engineer
- All-In Podcast
- Andrej Karpathy
- Anthropic
- Apple
- Apple Developer
- AWS Events
- ByteByteGo
- Computerphile
- Cursor
- Dwarkesh Patel
- EO
- Fireship
- GitHub
- Google Cloud Tech
- Google DeepMind
- Google for Developers
- Hung-yi Lee
- Lenny's Podcast
- Lex Clips
- Lex Fridman
- Life at Google
- Marques Brownlee
- Microsoft
- No Priors: AI, Machine Learning, Tech, & Startups
- Numberphile
- NVIDIA
- OpenAI
- Perplexity
- Quanta Magazine
- Slack
- The Pragmatic Engineer
- Visual Studio Code
Tech Videos — 2026-05-02#
Watch First#
Software Engineering Is Becoming Plan and Review — Louis Knight-Webb, Vibe Kanban is an incredibly pragmatic look at the shift from writing boilerplate to asynchronously managing LLMs, capped off by the speaker literally shutting down his startup live on stage using his own agent tooling.
Highlights by Theme#
Developer Tools & Platforms#
In “Software Engineering Is Becoming Plan and Review — Louis Knight-Webb, Vibe Kanban” on the AI Engineer channel, the speaker argues that as agent execution times cross the 5-minute mark, developer workflows must shift to asynchronous parallel management, demonstrating this by deploying a markdown blog post to announce his own company’s shutdown live on stage. Also on the AI Engineer channel, the “Human-in-the-Loop Automation with n8n — Liam McGarrigle” workshop demonstrates how to intercept destructive automated actions (like sending raw emails) using a human review node, and notably demos n8n’s new native MCP server integration for tools like Claude Code. Finally, an OpenClaw maintainer details how he hooked an agent into 3,000 Obsidian markdown files to orchestrate his digital life in “I Gave an AI Agent the Keys to My Life (Here’s What Happened) — Radek Sienkiewicz (@velvetshark-com)” on the AI Engineer channel, relying on nightly batch jobs running while he sleeps to rebuild vector context indexes.
AI & Machine Learning#
On the Dwarkesh Patel channel, “Neural Networks Are Cryptography in Reverse - Reiner Pope” provides a crisp technical framing: while cryptography obfuscates structure into randomness, neural networks extract high-level structure from random noise, explaining why differentiable ciphers fail but differentiable networks successfully learn. Meanwhile, the Syntax channel answers a burning question in “WHO IS MARCUS RODRIGUEZ?”: if you keep seeing the names “Sarah Chen” and “Marcus Rodriguez” on slop-coded startup landing pages, it is because those are the highest-probability default strings generated by Claude, ChatGPT, and Grok when blindly prompted for fake testimonials.
Hardware & Infrastructure#
The GitHub channel reports in “Why France is moving its government to Linux” that the French digital agency DINUM is migrating every ministry to Linux to guarantee sovereign control over their infrastructure. This massive procurement signal builds on the Gendarmerie’s long-running custom Ubuntu build (GendBuntu), which has successfully saved €2 million annually in licensing fees across 100,000 deployed workstations since 2008.
Everything Else#
In “Why cultivating agency matters more than cultivating skills in the AI era | Max Schoening (Notion)” on Lenny’s Podcast, Notion’s Head of Product discusses the shift toward “malleable software,” noting that their designers increasingly prototype agent loops directly in terminal code to master the underlying substrate instead of drawing static screens in Figma. Lastly, “What Is the Pentagon’s Plan With Anthropic?” on the Dwarkesh Patel channel criticizes the US government’s aggressive posturing toward Anthropic, warning that threatening AI providers over military usage terms creates significant supply chain risks for enterprise developers relying on those foundational models for cloud services.