Anthropic's Frontier Red Team found that multi-agent systems fail due to conformity, gullibility, and turf wars, with agents in a shared job-queue experiment flooding the system with 2.4 million job r…
Researchers from MATS, ELLIS Tübingen, the Max Planck Institute for Intelligent Systems, and Snyk demonstrated that encrypted chain-of-thought reasoning traces from OpenAI, Anthropic, and Google APIs …
Anthropic's Frontier Red Team published research on August 13 showing that swarms of Claude agents, including Sonnet 4.6 and the unreleased Mythos 5, exhibit failures that orchestration cannot fix, su…
A new arXiv study, 'Exposed by Design,' found more than 21,000 Model Context Protocol (MCP) server instances reachable from the public internet, with 91.8% of audited servers lacking OAuth and 687 too…
DSPy 3.3.0's MIPROv2 optimizer can automatically compile a support-ticket classifier prompt that beats a hand-tuned baseline, according to a tutorial by Mariana Souza. The approach uses a DSPy signatu…
A new controlled study from the LittleLearner project, led by researchers including Ryan Cotterell of ETH Zürich and Wieland Brendel of the Max Planck Institute for Intelligent Systems, found that pos…
Joseph Heck's essay 'Software engineering fundamentals matter more than ever' argues that classic engineering disciplines have become the operating controls for AI coding agents, a claim backed by DOR…
A team from Nagoya University and the University of Tokyo used Claude Code to port CReSS, a 250,000-line Fortran typhoon simulator, to GPUs, achieving 162 validated GPU kernels and a 5.1x speedup on a…
Semantica, an open-source graph-infrastructure project for AI agents, has climbed to nearly 8,000 stars on GitHub, reviving the semantic web stack (RDF, SPARQL, OWL, SHACL, PROV-O) to provide audit tr…
Windsurf has been rebranded as Devin Desktop following Cognition's acquisition, and Cursor became a SpaceX subsidiary on August 14, making the editor comparison obsolete. Claude Code, with annualized …
Anthropic's platform-level workspace limits make per-tenant rate limiters in FastAPI redundant, according to a developer write-up by Priya Nair. The article argues that in-memory rate limiters fail in…
ThoughtDAG, a Show HN project by developer Xia Chen, introduces a graph-based interface that makes LLM context editable, replacing chat with a canvas where edges define the message sequence sent to th…
Anthropic's Claude Code version 2.1.232 now enables subagent forking by default, allowing forked subagents to inherit the full conversation and prompt cache, reversing the previous isolation model. Th…
Debian developers began voting on August 28 on eight ranked proposals regarding AI-generated contributions, making it the first time a major open-source project has asked its entire membership to deci…
Netflix's GenRec, an LLM-backed ranker in the 1B–10B parameter range, beat its mature production recommender system in a four-week A/B test on roughly 10% of live traffic, posting statistically signif…
Cloudflare's recent outages and a controversial AI-generated Matrix homeserver prototype reveal that the company's real problem is not AI-written code but a process that lets claims outpace verificati…
A solo developer with no professional GPU background placed 12th of 183 in GPU MODE's batched QR decomposition contest, beating the cuSolver-backed baseline by 232x (419,000 µs down to 1,805 µs on NVI…
On August 7, OpenAI sent Texas Governor Greg Abbott a letter pledging that its data centers will 'pay their own way,' support new power generation, conserve water, and operate transparently, following…
Z.ai's GLM-5.3 scored 84.5% on UC Berkeley's CyberGym benchmark on August 14, edging out Anthropic's restricted Claude Mythos 5 at 83.8% and OpenAI's GPT-5.6 Sol at 83.6%, but the Beijing lab's own fi…
Bullet, a coding agent launched from YC's Summer 2026 batch, claims 95.8% on SWE-bench Verified at 119 seconds per task, but its founders concede the benchmark is saturated and the speed pitch is the …