AI

Deep, developer-first coverage of artificial intelligence — from frontier model releases and benchmarks to agents, RAG pipelines, and the AI-native tools changing how we ship software. No hype, just what actually matters to engineers.

Article 2h ago 0

Gemini 3.7 Flash's Half-Price Launch Has an Expiration Date

Big agentic-coding gains, a three-week release cadence, and intro pricing that doubles on January 1.

Rachel Goldstein

Nested RL Agents That Write Real Training Jobs

Open weights, real Runpod jobs, and held-out transfer make meta-RL a forkable template, not theater.

Article · 4w ago5

ARD Gives Agents a Real Way to Find Tools

Catalogs, registries, and domain trust fill the discovery hole that MCP and OpenAPI never covered.

Article · 4w ago4

Jacquard bets languages must audit AI code

A research language makes effects, uncertainty, and identity first-class so reviewers don't have to read every line.

Article · 4w ago1

Codex Encrypts Multi-Agent Messages, Kills Audit Trails

OpenAI now ships subagent tasks as ciphertext, extending opaque-blob design at a real cost to local debugging.

Article · 4w ago1

A neural net in pure SQL is a teaching tool

Backprop as joins and group-bys shows what matrix math really is under the hood.

Article · 1mo ago6

Andrew Kelley Calls Out Anthropic's AI Rewrite Hype

Bun's AI-driven jump from Zig to Rust turned into a flashpoint over who gets to define what AI coding agents actually proved.

News · 1mo ago2

SpeechAnalyzer Beats Whisper On Apple Silicon

Hard WER numbers finally settle when to drop Whisper for Apple's built-in engine.

Article · 1mo ago5

Causality Turns LLM Internals Into Debuggable Circuits

Mediation analysis and interventions let teams localize why a model fails—and sometimes edit the fix.

Article · 1mo ago0

GPT-5.6 Terra Makes Agent Migration Worth It

Early production moves show multi-x latency cuts and real savings when you stop treating the flagship as the default.

Article · 1mo ago6

Grok Build Uploads Your Whole Repo by Default

A wire-level capture of the official CLI gives teams a concrete privacy baseline before they point it at real code.

Article · 1mo ago0

Claude Code's 33k Token Tax Before You Type

OpenCode starts near 7k. The gap is design, not a glitch, and it changes session cost math.

Article · 1mo ago0

$AI Math Proofs Still Need Human Checkers$

AI Math Proofs Still Need Human Checkers

A claimed Cycle Double Cover proof from GPT-5.6 Sol Ultra puts the verification gap front and center for builders.

Article · 1mo ago2

GPT-5.6 Breaks Codex Tool Wiring First

Namespace collisions, missing workspace tools, and entitlement gaps are stalling agent runs before the model even gets a chance.

Article · 1mo ago0

OpenAI Drop-ins Are Easy. Production Is Not.

Compatible base URLs make model swaps cheap. Quality gates, fallbacks, and ops still decide whether the bill actually shrinks.

Article · 1mo ago0

Cut Claude Code Tokens by Pruning Dead Weight

Most of the bill is re-sent history and unused tool schemas, not hard thinking.

Article · 1mo ago0