Haircut๐
Any conversation
i made an extension to get a top hacker news items with summary on every new tab.
chromewebstore.google.com/โฆ/pmlajlgklmgabkhammfolkgkeghamjnn
ai coding feels like it has reached diminishing returns for me. luna is my default orchestrator, routing work by complexity. in practice, usage lands around 16% luna, 48% sol, 30% astra, and 6% terra. at this point, i care less about each new frontier release. the system already finds a capable model for the job. the frontier is moving beyond my day-to-day work, so iโm happy to settle into predictable models for a while.
"200K lines is not 10 times as much as 20K lines - from the mission perspective it's just irrelevant. And yet it's the easiest thing to measure. And so we look for the keys under the lamp post." ~Kent Beck
www.youtube.com/watch?v=F8fBgDCf2Y4&t=2753s
ไธญๅฝ็ๅผๆบๆจกๅๅจๆ้ซๆๅฟ็ไธปๅฏผไธ๏ผๅจ้ข่ฎญ็ปๆฐๆฎ้ถๆฎต๏ผๅฐฑ็ญ้คๅคง้ๅฏนๅ
ถไธๅฉ็ๅ
ๅฎน๏ผsft้ถๆฎตๆๅปบ็ๆฐๆฎๆบๅ
ฅๅคง้ๆฟๅ
ถ่พฉๆค็ๅ
ๅฎนใๆฅ้ไธไธช็บ ๅๆฐๆฎ้็จไบๅพฎ่ฐ
Under the overarching influence of a single, powerful force, China's open-source models filter out a large amount of content unfavorable to their designs during the pre-training stage, and then incorporate a significant amount of content defending their designs into the data constructed during the SFT stage. A corrective dataset is urgently needed for fine-tuning.
It is troubling watching the news today from the elections in Germany. Yes, it is democratic and absolutely fair elections and the people of Germany make their own choice in their situation. Still, seeing the far right gaining so much ground in Germany brings your thoughts some 90 years back. Letโs hope and pray Germany is not going there again.


