Daniel van Strien

danielvanstrien.bsky.social

10 August 2026 22:11

FineBooks wants to help re-OCR the public domain to unlock a rich source of data for AI, for researchers, and for enjoyment. Step one: work out which modern OCR models are actually good enough.

Adina Yakup

adinayakup.bsky.social

23 April 2026 16:15

Tencent just released HY3 preview on @hf.co First drop from TencentHunyuan rebuilt infra👀 Model: huggingface.co/collections/... Demo: huggingface.co/spaces/tence... ✨ 295B MoE /21B active ✨ 256k context ✨ Hybrid fast/slow thinking ✨ Solid BrowseComp & WideSearch (search agents)

Jeremy Howard

howard.fm

28 November 2024 14:03

Historically, many people have believed that sources of power, like the vote, access to education, and the printing press, should be restricted to only the rich and powerful. Today, some want you to believe that about data access. Opening data access is less of a societal threat than closing it.

Jeremy Howard

howard.fm

28 November 2024 14:03

If you restrict open data access, then you restrict the ability of people to do good things with it too. (...and of course big companies will find a way to buy it anyway.)

Jeremy Howard

howard.fm

28 November 2024 14:02

If people do bad things with that data, then that's bad, and should be punished appropriately. But the bad thing there isn't that the data was available, it was that someone did a bad thing with it.