Models & Releases
New AI model releases, benchmarks, comparisons, and open source drops
20 articles
Microsoft patches Windows and Excel – breaks audio, remote access, and paste
Show HN: Neobrutalism.dev – Just added Base UI support and added new color theme
Distributed Systems Classics
Truncated SVD (2023)
Models & Releases
UkisAI Swift-Qwen3.8-27B / -58.3% thinking, x1.95 speed while keeping the accuracy of xhigh
Hi everybody, we post-trained Qwen 3.8 27B to be more efficient by figuring out which tokens were linked to overthinking and penalizing them without "attacking" the reasoning length directly then fixed the accuracy with a bit of secret s...
Are there any organizations that are lobbying in favor of open source AI?
So we’re seeing how Anthropic and OpenAI are gunning for regulations. I think most of us realize that this is a ploy for them to achieve regulatory capture, thus securing their moat and kicking out open source. The thing is, there’s so ...
A Beginning for Mathematics
China's Regulators Take Aim at "AI Boyfriends"
Principles for Fast Tokio Applications
Foundation Model Engineering: From Theory to Production
Things That Annoy Me About Cars
Looking for advice for small office looking for local AI RAG
My company is looking for basically a local hardware back-end for an already set up Open WebUI Windows AD joined setup. We are already on the frontier models thru Open WebUI. ~10 users. Workload is a lot of contract and chat based email...
RubyGems Open Source Supply Chain Security and OpenAI
Jabber/XMPP: How Do We Gain Traction?
Models & Releases
Xi promotes open source AI zone among BRICS countries
Xi emphasised the need to step up cooperation. https://www.hindustantimes.com/india-news/china-will-lead-creation-of-open-source-ai-for-brics-xi-jinping-101789300908632.html https://www.msn.com/en-xl/news/other/xi-jinping-proposes-ai-ope...
New $100K H-1B Visa Fee Pushes Tech Jobs Offshore
Adversarial Fashion Makes a Statement on AI Panopticon
Models & Releases
K2 Horizon lineup is out on AA, and once again AA plots are misleading.
The full K2 Horizon lineup is out on Artificial Analysis. The AA intelligence vs. parameters plots show that - 0.9B and 375B are bad - 3.7B and 7B are SOTA - 36B A4B is SOTA for hardware with poor memory bandwidth (spilled experts, Strix...