What I found interesting this time

“The Generative AI Learning Penalty: Evidence from Chinese Secondary Education,” n.d. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6868618.

“GitHub Has Alternatives, but No Replacement,” n.d. https://lalitm.com/post/github-alternatives/

Valid critiqe of various alternatives to github. I think I should really write my own blog post about it.

“AI;DR (AI; Didn’t Read),” n.d. https://www.rickmanelius.com/p/aidr-ai-didnt-read

People stop reading AI articles and type AI;DR. I love it!
If you have not spent time writing text why would I bother to read it.

“Alibaba’s TSMC-Built 5nm RISC-V Chip, XuanTie C950, Now Runs Qwen-3.8 27B Model Natively, Unlocking Massive Vertical Integration Tailwinds,” n.d. https://wccftech.com/alibabas-tsmc-built-5nm-risc-v-chip-xuantie-c950-now-runs-qwen-3-8-27b-model-natively-unlocking-massive-vertical-integration-tailwinds/

I mention this before but will also repeat it again. China has nothing to loose when it comes to AI advancements. USA is a different story. If someone creates good local LLM their economy will go to shit.
And it looks like with right optimizations for which US companies dont have time to invest for you can literally run huge LLM right on your own hardware.

“5.0: Laying the Foundation,” n.d. https://blog.joinmastodon.org/2026/08/5.0-laying-the-foundation/

They are finally fizing vocabulary of Mastodon. No more lists or private mentions. Good news!

“AI Can Find Zero-Days but Still Can’t Reliably Write Secure Code,” n.d. https://www.csoonline.com/article/4210735/ai-can-find-zero-days-but-still-cant-reliably-write-secure-code.html.

“How Claude Marks AI-Generated Content,” n.d. https://support.claude.com/en/articles/16266773-how-claude-marks-ai-generated-content.

“Jak Odpowiedzi AI Będą Znakowane w Sposób Niewidzialny Dla Ludzi?,” n.d. https://niebezpiecznik.pl/post/jak-odpowiedzi-ai-beda-znakowane-w-sposob-niewidzialny-dla-ludzi/.