What’s new. What matters. What do you think?
OpenAI published an Australian youth-safety blueprint as a government consultation nears its deadline. Here is what is shipped, proposed and not yet law.
Sign in to comment or vote. Sign in
Open the full discussionAWS says Kiro Crew helped merge 1,000 pull requests in seven days. The reusable lesson is its scaling ladder, permission model, audit trail and cost boundary.
Sign in to comment or vote. Sign in
Loading comments…
Step 5 Preview now has inspectable weights and serving configurations. This update corrects the earlier release forecast and explains the StepFun license boundary.
Sign in to comment or vote. Sign in
Loading comments…
Grok Voice Transcribe 2.0 adds batch and streaming transcription, diarization, eight channels and low hourly pricing. One version discrepancy matters.
Sign in to comment or vote. Sign in
Loading comments…
SGLang 0.5.20 adds eight model families, changes Responses API storage to opt-in, removes prefill CP v1 and retires CUDA 12 artifacts.
Sign in to comment or vote. Sign in
Loading comments…
Qwen Code 0.24.1 adds Playwright browser control, live voice, containerized subagents and workflow checkpoints. Here is the safe rollout map.
Sign in to comment or vote. Sign in
Loading comments…
Qwen3.8-Omni-Flash accepts text, image, audio and video in a 1M-token context. Here is what ships, what does not, and what teams should test.
Sign in to comment or vote. Sign in
Loading comments…
The SPARSEUP sparse retriever maps queries and documents to weighted vocabulary terms. See its 149M design, benchmarks, failure modes and RAG test plan.
Sign in to comment or vote. Sign in
Loading comments…
The Unity agent plugin adds official Unity 6 workflow guidance to Claude Code and Codex. Learn installation, verification, license and safe testing steps.
Sign in to comment or vote. Sign in
Loading comments…
Claude Code AGENTS.md support adds a fallback for project instructions, while version 2.1.278 moves eligible auto mode checks to an unbilled server path.
Sign in to comment or vote. Sign in
Loading comments…
Google MilleMiglia generates privacy-preserving middle-mile logistics instances with schedules, capacity and synchronization constraints for reproducible research.
Sign in to comment or vote. Sign in
Loading comments…
Tencent EVIE open-sources 4.5B and 8B visual document retrieval models. Learn how Prefix-MRL and HAC trade index size against retrieval quality.
Sign in to comment or vote. Sign in
Loading comments…
Loading comments…