Hackernews posts about Mox

1. What happens when a kid's robot best friend dies? (www.technologyreview.com)
   4 points by Brajeshwar 4 days ago | discuss
2. 3D Visualization of E8 Using an H4 Folding Matrix [pdf] (www.researchgate.net)
   1 points by brandonb 18 days ago | discuss
3. Muse Glimmer: 30B-parameter model optimized for always-on local agent workflows (research.meta.ai)
   1208 points by riordan 11 days ago | 639 comments
4. Our position on open-weights models (www.anthropic.com)
   1180 points by surprisetalk 24 days ago | 1747 comments
5. Qwen3.8-Max: A New Bar for Coding and Cowork (qwen.ai)
   1124 points by ai2027 18 days ago | 615 comments
6. AMD acquires Taalas to boost inference performance by etching models in silicon (www.theregister.com)
   949 points by itvision 14 days ago | 723 comments
7. Windows 11's built-in Weather app wastes more than 1 GB of RAM (www.notebookcheck.net)
   671 points by akyuu 12 days ago | 598 comments
8. Memory prices climb 500% in 12 months (www.tomshardware.com)
   665 points by haunter 3 days ago | 560 comments
9. Nvidia, Microsoft, Meta warn against overregulating open-weight models (www.cnbc.com)
   659 points by louiereederson 28 days ago | 318 comments
10. More German than many Germans (mertbulan.com)
    647 points by mertbio 18 days ago | 500 comments
11. Mark Zuckerberg attacks 'closed' AI rivals as Meta returns to open models (www.ft.com)
    642 points by root-parent 11 days ago | 601 comments
12. GigaToken: ~1000x faster Language model tokenization (github.com)
    620 points by syrusakbary 29 days ago | 120 comments
13. Moderna reports first positive Phase 3 for mRNA neoantigen therapy in melanoma (twitter.com)
    589 points by heydenberk 1 day ago | 286 comments
14. Google fixed more Chrome bugs in June than over the past two years, thanks to AI (blog.google)
    585 points by Garbage 21 days ago | 610 comments
15. Qwen3.8 Max now ranked as the best overall model by agentic index (artificialanalysis.ai)
    546 points by apitman 14 days ago | 354 comments
16. Show HN: I trained a 125M model to autocomplete piano on-device (simedw.com)
    530 points by simedw about 15 hours ago | 110 comments
17. Show HN: Echo – Fable-level results at 1/3 the cost using open-weight models
    484 points by adam_rida 28 days ago | 228 comments
18. Mistral's Shieldstral: 3B open-weights model for multimodal moderation (mistral.ai)
    482 points by riadsila 16 days ago | 133 comments
19. Controversial creators are benefiting from monetization programs run by Meta (www.abc.net.au)
    479 points by robtherobber 9 days ago | 338 comments
20. The new rules of context engineering for Claude 5 generation models (claude.com)
    463 points by mellosouls 26 days ago | 404 comments
21. DeepMind's WeatherNext model achieves breakthrough forecasting cyclones (deepmind.google)
    449 points by bhavansig 13 days ago | 130 comments
22. Beating GPT-5.6 Sol on retrieval with 100x cheaper open models (neon.com)
    437 points by moonikakiss 15 days ago | 127 comments
23. Mojo 1.0 (www.modular.com)
    430 points by dayanruben 9 days ago | 244 comments
24. Open-weight AI is having its Kubernetes moment (tobi.knaup.me)
    416 points by tknaup 27 days ago | 322 comments
25. More Tailscale tricks for your jailbroken Kindle (tailscale.com)
    410 points by Error6571 23 days ago | 115 comments
26. A missing underscore sent innocent man to prison for 18 months (arstechnica.com)
    408 points by quantified 24 days ago | 293 comments
27. Apple says more ex-employees may have taken confidential data to OpenAI (techcrunch.com)
    389 points by thewebguyd 16 days ago | 285 comments
28. Nine PBS sues Iron Mountain over blocked access to archival data (current.org)
    386 points by vinayakborkar 8 days ago | 226 comments
29. Demis Hassabis is moving from CEO to Chairman at Google DeepMind (www.axios.com)
    371 points by ot 15 days ago | 1 comments
30. GPT 5.6 Sol is the best "vision" model OpenAI ever released (blog.roboflow.com)
    365 points by plurby 4 days ago | 170 comments