more tekacs's favorites

31.		LLMs can see and hear without any training (github.com/facebookresearch)
		210 points by T-A 11 days ago \| 66 comments
32.		Lossless LLM compression for efficient GPU inference via dynamic-length float (arxiv.org)
		411 points by CharlesW 12 days ago \| 117 comments
33.		Reproducibility project fails to validate dozens of biomedical studies (nature.com)
		267 points by rntn 12 days ago \| 125 comments
34.		Avoiding skill atrophy in the age of AI (addyo.substack.com)
		373 points by NotInOurNames 12 days ago \| 317 comments
35.		A $20k American-made electric pickup with no paint, no stereo, no screen (theverge.com)
		1457 points by kwindla 12 days ago \| 1297 comments
36.		Show HN: Index – New Open Source browser agent (github.com/lmnr-ai)
		98 points by skull8888888 14 days ago \| 45 comments
37.		Launch HN: Cua (YC X25) – Open-Source Docker Container for Computer-Use Agents (github.com/trycua)
		172 points by frabonacci 14 days ago \| 73 comments
38.		LLM-powered tools amplify developer capabilities rather than replacing them (matthewsinclair.com)
		345 points by matthewsinclair 16 days ago \| 238 comments
39.		Ultra-precision formation flying demonstration for space-based interferometry (arxiv.org)
		70 points by PaulHoule 16 days ago \| 23 comments
40.		The Future of Compute: Nvidia's Crown Is Slipping (mohitdagarwal.substack.com)
		144 points by wilson090 16 days ago \| 120 comments
41.		Show HN: Open Codex – OpenAI Codex CLI with open-source LLMs (github.com/codingmoh)
		106 points by codingmoh 16 days ago \| 37 comments
42.		An update to our pricing (windsurf.com)
		127 points by mefengl 16 days ago \| 92 comments
43.		Gemini 2.5 Flash (googleblog.com)
		1076 points by meetpateltech 20 days ago \| 560 comments
44.		Show HN: Plandex v2 – open source AI coding agent for large projects and tasks (github.com/plandex-ai)
		257 points by danenania 21 days ago \| 81 comments
45.		Microsoft researchers developed a hyper-efficient AI model that can run on CPUs (techcrunch.com)
		146 points by libpcap 21 days ago \| 64 comments
46.		Cursor IDE support hallucinates lockout policy, causes user cancellations (reddit.com)
		1511 points by scaredpelican 23 days ago \| 606 comments
47.		Generate videos in Gemini and Whisk with Veo 2 (blog.google)
		347 points by meetpateltech 22 days ago \| 132 comments
48.		Skywork-OR1: new SOTA 32B thinking model with open weight (github.com/skyworkai)
		145 points by naomiclarkson 24 days ago \| 28 comments
49.		Google will let companies run Gemini models in their own data centers (cnbc.com)
		509 points by jonbaer 28 days ago \| 182 comments
50.		Google is winning on every AI front (thealgorithmicbridge.com)
		1005 points by vinhnx 26 days ago \| 819 comments
51.		Wasting Inferences with Aider (worksonmymachine.substack.com)
		139 points by Stwerner 24 days ago \| 105 comments
52.		Colossus for Rapid Storage (cloud.google.com)
		244 points by alobrah 28 days ago \| 117 comments
53.		SpacetimeDB (spacetimedb.com)
		347 points by matthewfcarlson 28 days ago \| 176 comments
54.		UK Effort to Keep Apple Encryption Fight Secret Is Blocked (msn.com)
		171 points by flanked-evergl 29 days ago \| 139 comments
55.		Tailscale has raised $160M (tailscale.com)
		659 points by louis-paul 29 days ago \| 332 comments
56.		Apache ECharts (apache.org)
		1658 points by tomtomistaken 29 days ago \| 250 comments
57.		Neural Graffiti – Liquid Memory Layer for LLMs (github.com/babycommando)
		107 points by vessenes 30 days ago \| 25 comments
58.		Framework temporarily pausing some laptop sales in the US due to tariffs (fosstodon.org)
		194 points by leotravis10 30 days ago \| 44 comments
59.		Apple’s Darwin OS and XNU Kernel Deep Dive (tansanrao.com)
		485 points by tansanrao 32 days ago \| 180 comments
60.		The Llama 4 herd (meta.com)
		1235 points by georgehill 32 days ago \| 658 comments
		More