Open-Source AI
Open models — Llama, Mistral, DeepSeek, Qwen and others — are catching up to closed labs and let you run AI on your own hardware, vendor-free. That shifts the balance of power, the safety debate and the economics of the field. Here is everything on open-source AI: open-weight releases, licenses, local deployment and the race against closed models.
Alibaba unveils $450 AI smart glasses and expands its gadget lineup
Alibaba unveiled a new generation of smart glasses priced at about $450. The company keeps moving its AI from the cloud into consumer devices, challenging Meta and…
RAG Failures: 6 Reasons Your AI Agent's Memory Lags, and How to Diagnose Retrieval
73% of RAG system breakdowns are caused by bad search. SpeShuAI breaks down 6 typical failure modes — from chunking that destroys context to semantic gaps between…
Atlassian Does Not Use Chinese AI Models in Production, CEO Cannon-Brookes Says
Atlassian CEO Mike Cannon-Brookes confirmed in a 20VC podcast interview: the company does not use Chinese AI models in production despite growing market pressure.
Meta releases Muse Wonder and claims a big leap over its predecessors
Sourcegraph's Amp Learns to Watch Its Own PRs: Comments, Reviews, and CI in a Single Thread
Meta Shares Jump on New Muse Wonder Model as Market Waits for Numbers
Meta releases Muse Wonder: shares rise, but the numbers are still under wraps
Can AI reproduce itself? Anthropic ran an experiment on 15 cheap streaming boxes
Anthropic spent $825 on 15 Android TV boxes to test whether a neural network could independently replicate itself across a network. The results showed that the…
Meta Releases Muse Wonder and Wonder Full — Rivals to GPT-6 Astra and Claude Fable
Meta has released two models, Muse Wonder (54.3) and Wonder Full (59.1 points), to compete with OpenAI and Anthropic. They beat GPT-6 Astra at coding: in DeepSWE…
Developer rewrites seven Adobe apps in Rust and open-sources the code
OpenAI GPT-6 Sol, Luna and Astra Now Available in Bedrock From Five Asian Regions
OpenAI Expands Its Dublin Headquarters: New Jobs at the EU Office
11 DevOps Tools: A Buyer's Guide to Every Layer of Your Stack
ByteDance Researchers Find a Weak Spot in DeepSeek's Memory-Saving Approach
A paper from ByteDance with co-authors from Princeton, Stanford and Berkeley questions the accuracy of DeepSeek models built on a memory-saving technique.
An AI secretary for $77 a month: 88% of answers from DeepSeek, 77% of the budget goes to Opus
Telegram bot Liza Petrovna costs $77 a month: 88% of replies run on cheap DeepSeek ($0.002-0.007), while the 7% of hard queries routed to Opus ($0.21-0.24) eat up…