Techmeme

Harvard releases Institutional Books 1.0, a dataset for AI researchers with 242B tokens, from 394M scanned pages and 983K public domain books in 254 languages (Matt O'Brien/Associated Press)

Matt O'Brien / Associated Press: Harvard releases Institutional Books 1.0, a dataset for AI researchers with 242B tokens, from 394M scanned pages and 983K public domain books in 254 languages  —  Everything ever said on the internet was just the start of teaching artificial intelligence about humanity.

Jun 14, 2025 - 05:30
 0
Harvard releases Institutional Books 1.0, a dataset for AI researchers with 242B tokens, from 394M scanned pages and 983K public domain books in 254 languages (Matt O'Brien/Associated Press)

Matt O'Brien / Associated Press:
Harvard releases Institutional Books 1.0, a dataset for AI researchers with 242B tokens, from 394M scanned pages and 983K public domain books in 254 languages  —  Everything ever said on the internet was just the start of teaching artificial intelligence about humanity.

Read More

Tags:

Previous Article

Why is digital sovereignty important right now?

Next Article

Canary, which offers hotel guest management tools, raised an $80M Series D led b...

Related Posts

Source: ~1,000 people have left CISA during the second Trump administration, including 750+ taking a buyout, cutting CISA's workforce by nearly a third (Sam Sabin/Axios)

Source: ~1,000 people have left CISA during the second ...

Jun 4, 2025  0

The Claude 4 models support "extended thinking with tool use", a beta feature that lets them alternate between reasoning and using tools like web search (Benj Edwards/Ars Technica)

The Claude 4 models support "extended thinking with too...

May 22, 2025  0

Texas enacts an online child safety bill requiring app stores to verify users' ages and secure parental approval before minors can get most apps or make IAPs (Bloomberg)

Texas enacts an online child safety bill requiring app ...

May 28, 2025  0