These AI chatbot medical advice fails are apropos of my recent post: danielsolove.substack.com/p/ai-compani... "Doctors and therapists have years of training and experience . . . and well-established legal responsibilities. . . AI currently has none of these things."

The primary problem of AI isn't AI; the problem is the wrong people are in charge of building it.

Imagine if we could have AI developed not by a bunch of billionaires whose brains were fried by misreading sci-fi but by a group more representative of society, a group that was thoughtful and responsible and concerned about the impact. Then, AI might actually be a force for good.

AI isn't the enemy. It's the ethos of those who build it that's toxic. Far too often, policymakers hail it as "innovation" but it's not. The business models are based on taking the data and IP of others and on externalizing costs and risks.

AI has been rolled out like most digital technologies. No attempt to seek input from a diverse group of stakeholders. No care to listen to concerns about the impact. Steal data online, greedily gobble up resources (driving up costs for everyone), grow as fast as possible no matter the harms.

7. The project of @daniel-solove.bsky.social's latest book, On Privacy and Technology, is to synthesize twenty five years of thinking about privacy into a “succinct and accessible” volume and to help the reader understand “the relationship between law, technology, and privacy.”