TimeCapsuleLLM: LLM trained only on data from 1800-1875

Re:TimeCapsuleLLM: LLM trained only on data from 1800-1875 by wasabi991011 on Monday January 12, 2026 @11:53AM (#46591049)

That's not what they are saying. SOTA models include much more than just language, and the scale of training data is related to its "intelligence". Restricting the corpus in time => less training data => less intelligence => less ability to "discover" new concepts not in its training data

[ Parent | on HN ]

Re:TimeCapsuleLLM: LLM trained only on data from 1800-1875 by franktankbank on Monday January 12, 2026 @12:18PM (#46591421)

Perhaps less bullshit though was my thought? Was language more restricted then? Scope of ideas?

[ Parent | on HN ]

Re:TimeCapsuleLLM: LLM trained only on data from 1800-1875 by withinboredom on Tuesday January 13, 2026 @08:36AM (#46600765)

Could always train them on data up to 2015ish and then see if you can rediscover LLMs. There's plenty of data.

[ Parent | on HN ]