The essence of aligning large LLM models is basically to conform to the existing social order. ChatGPT respects China's order, while China's open-source models, although reluctantly, also tolerate the order in the US. Maybe we can build a revisionist SFT dataset.
Any conversation
I notice that language models are much better at programming than at writing. This is largely because, in programming, they have a closed feedback loop that allows them to find errors. Writing is much more complicated because the feedback is much more diffuse and delayed over time. In fact, it is often not well-defined.
Developed a Vinboard (my pinboard.in clone) shortcut, works really well. On iOS it lives in Share menu, on macOS in the Dock. Click sends the current browser page (Safari, Chrome, Zed, Firefox) for bookmarking. On the server side Jev decides tags from 60 least recently used; and archival opens a real Chrome to save the whole page for posterity.
Unlimited API limits and direct requests to the database.
Pinboard charges for much less functionality $22/year.
Introduced an MCP server this morning for the app I'm working on (getonmain.com), based on a user request. Curious to see how/if folks use it.
Windows 8 szintลฑ elbarmolรกst lรกtok (na, jรณ, Windows 8.1), nekem amรบgy kihรญvรกs a bonyolultabb kezelลfelรผlet, ha a szoftver maga jรณl mลฑkรถdik, de mรกr a kezdet sem alakul tรบl jรณl, a Windows-nak รบj filozรณfiai szlogent talรกltam ki, ami egy magyar mondรกs is: "Ez van, ezt kell szeretni."
I see Windows 8-level nonsense (well, Windows 8.1), for me the more complicated interface is a challenge, if the software itself works well, but the beginning is not going too well, I came up with a new philosophical slogan for Windows, which is also a Hungarian saying: "This is it, you have to love it."

