rollup.
293 writings found
Page 5
LLM 0.32a0: Why Your Prompt/Response Mental Model is Already Obsolete
Simon Willison's LLM library gets a major refactor to handle multi-modal inputs, streaming message parts, and the messy reality of modern AI models
The Fundamental Shift: Why Verification Speed Matters More Than Code Generation
AI coding tools are everywhere now. But the game isn't about generating code faster anymore. It's about knowing if that code is right.
LLM 0.32a0: When Your Abstraction Meets Reality's Complexity
Simon Willison's LLM library gets a major refactor to handle the messy, multi-modal world of modern AI models. Here's why abstractions always break.
Meta's HSM Vault: Why Cryptographic Transparency Matters More Than the Encryption Itself
Meta's HSM-based backup vault gets over-the-air key distribution and public deployment evidence. The real story is about verifiable trust, not just encryption.
Meta's HSM Vault: The Infrastructure Behind Encrypted Backups You Actually Can't Access
Meta's publishing cryptographic proof of their HSM deployments. Here's why this matters for encrypted backups and what developers should know.
GitHub Copilot CLI: Why Two Modes Matter More Than You Think
Interactive vs non-interactive in Copilot CLI isn't just about speed. It's about fundamentally different ways of thinking through problems at the terminal.
Facebook's Search Problem: Why Keywords Can't Find Cupcakes
Meta rebuilt Facebook Groups search from scratch, blending dense embeddings with inverted indices. Here's why that matters for community search.
Google's ERA is Quietly Proving AI Can Actually Do Science
Google Research's Empirical Research Assistance moves beyond proofs-of-concept into real epidemiology, cosmology, and neuroscience applications
Why I Actually Care About Markdown Now (And You Should Too)
Markdown isn't just formatting syntax. It's the quiet infrastructure of developer communication that makes or breaks your open source presence.
Google Photos' AI Wardrobe: Why Your Closet Doesn't Need Machine Learning
Google's new AI-powered digital closet feature sounds impressive, but the technical reality reveals deeper questions about consumer AI applications.
Treating Prompts Like Code: What SPDD Gets Right About AI-Assisted Development
Thoughtworks' Structured Prompt-Driven Development treats prompts as version-controlled artifacts. Here's why that matters more than typing faster.
Treating AI Prompts Like Code: What I Learned From Thoughtworks' SPDD Method
Thoughtworks turned AI coding assistants into team assets with Structured Prompt-Driven Development. Here's why versioned prompts might matter more than speed.
Microsoft and OpenAI Just Rewrote Their Relationship (And Everyone Wins)
The new Microsoft-OpenAI deal solves a brewing legal battle with Amazon, reshapes cloud exclusivity, and changes everything for enterprise AI deployment.
Permission Hungry Agents and the Return to First Principles
ThoughtWorks Radar 34 reveals AI's paradox: tools that generate complexity at speed while forcing us back to security basics and software fundamentals.
Project Maven: How AI Turned War Into a Database Query
The military's AI targeting system can now hit 5,000 targets daily. We should talk about what happens when warfare moves at database speed.