Hacker News: gatuamgbhttps://news.ycombinator.com/user?id=gatuamgbHacker News RSShttps://hnrss.org/hnrss v2.1.1Mon, 19 Oct 2026 18:47:07 +0000<![CDATA[Daily MCP Tool Drifts- 8,931 of 12,391 drifted with no version bump]]>Article URL: https://mcpindex.ai/ledger

Comments URL: https://news.ycombinator.com/item?id=49351487

Points: 1

# Comments: 0

]]>Tue, 18 Aug 2026 19:40:17 +0000https://mcpindex.ai/ledgergatuamgbhttps://news.ycombinator.com/item?id=49351487https://news.ycombinator.com/item?id=49351487<![CDATA[New comment by gatuamgb in "Show HN: Bento - An entire PowerPoint in one HTML file (edit+view+data+collab)"]]>It would be a fun exercise to test it with small language models- seeing how this works. If the results are good, there could be a good path to drive mass adoption of the concept!

]]>Thu, 23 Jul 2026 03:22:20 +0000https://news.ycombinator.com/item?id=49016505gatuamgbhttps://news.ycombinator.com/item?id=49016505https://news.ycombinator.com/item?id=49016505<![CDATA[New comment by gatuamgb in "Petals: Run LLMs at home, BitTorrent-style"]]>You can easily build a localized, access-controlled compute grid using Apple Silicon hardware. Haven't tested but as far as I can see, the key to making it performant would be keeping full models resident on individual nodes rather than attempting to distribute the tensor processing across the network.

]]>Thu, 23 Jul 2026 03:04:28 +0000https://news.ycombinator.com/item?id=49016361gatuamgbhttps://news.ycombinator.com/item?id=49016361https://news.ycombinator.com/item?id=49016361