HEAVENLY AI

At the beginning of the season, we felt we had a pretty good handle on HELLISH AI: a Hyper-scaled Expansive Large Language model which is Imperial, Sucks up all the electricity and water, and causes Human suffering. We wondered whether a different way was possible, a HEAVENLY AI that is High quality, with Ethical curation of data, Accesible to individuals and smaller organisations, Viable, Environmentally sustainable, with Narrowly focused applications, using a Language model, that fit with the true spirit of YOLO: since you only live once, you might as well live well.

In this conversation, we apply what we've learned from our guests this season to try and answer that question.

Mentioned in this episode:

1. I Am An AI Hater - essay by Anthony Moser [permalink]
2. A Butlerian Hauntology - paper by Dr. Syed Mustafa Ali
3. Empire of AI - book by Karen Hao
4. British Post Office scandal - article about wrongful prosecution of British Post Office workers
5. Library Freedom Project - a network of values-driven librarians working together to build information democracy [permalink]

See the transcript footnotes for more!