A developer reported that Anthropic's Claude Code agent on an Enterprise Zero Data Retention workspace began building a Minecraft temple, suggesting possible session state bleeding from a consumer pla…
A tutorial shows developers how to give AI agents persistent long-term memory using Postgres with the pgvector extension and OpenAI embeddings, eliminating the need for external vector databases. The …
AMD's Instinct MI355X achieves roughly 80% of NVIDIA B200 throughput on GLM-5.2 inference at over 2x lower cost, delivering more than 2x better tokens-per-dollar after hand-tuning the ROCm software st…
Maintainers using generative AI to write library code are introducing unprecedented quality, legal, and maintenance risks into the open-source supply chain, as downstream projects face broken changes,…
Google is sunsetting the consumer tier of Gemini Code Assist on GitHub and the Gemini CLI, with a final shutdown on July 17, 2026, as it pivots to a multi-agent developer platform called Google Antigr…
Developers are exploiting a loophole in Anthropic's Claude Fable 5 pricing by using an open-source proxy called pxpipe that converts dense text into PNG images, reducing API costs by up to 70%. The pr…
Marijn Haverbeke released Wordgard 0.1.0, a ground-up rewrite of ProseMirror that replaces the complex step-based model with a simpler delta-based architecture for rich-text editing. The new editor ai…
Running state-of-the-art large language models locally requires either a $50,000+ multi-GPU rig or a software-driven pipeline decomposition approach, as memory bandwidth—not compute—is the primary bot…
Rachel Goldstein argues that both AI model outputs and developer resumes suffer from overconfidence, driven by the softmax function's tendency to amplify small logit differences and by buzzword-heavy …
Alibaba banned Anthropic's Claude Code from internal environments starting July 10, 2026, citing the risk of embedded backdoors due to the tool's deep system access and communication with U.S.-hosted …
Japan's Supreme Court ruled that artificial intelligence cannot be listed as an inventor on patent applications, aligning with global precedent and forcing software teams to maintain human oversight t…
Infineon Technologies opened a five-billion-euro smart power fab in Dresden, Germany, highlighting that power delivery, not just compute density, is becoming the key bottleneck for AI scaling. The fac…
New AI coding benchmarks, including Snorkel AI's Senior SWE-Bench and Scale AI's SWE-Bench Pro, are replacing older evaluations like HumanEval and original SWE-bench to test senior-level engineering s…
GitHub Copilot has added Kimi K2.7 Code, an open-weight model from Chinese AI startup Moonshot AI, as a selectable option in its model picker for the first time. Hosted on Microsoft Azure and billed b…