This is how AI bias really happens — and why it’s so hard to fix | Curio
News and insights read out for you.
10,000+ audio articles. 50+ world-leading publications.

All in 1 subscription.

MIT Technology Review logo

This is how AI bias really happens — and why it’s so hard to fix

10 mins | Jun 14, 2020
story image

AI models we create are mirrored, consciously or unconsciously, after our own views. Bias can creep in at many stages of the deep-learning process, and the standard practices in computer science aren't designed to detect it. MIT Technology Review's Karen Hao has written this piece illuminating the many stages where bias can be transferred into AI models.

Get unlimited access free for 7 days, then $6.67/month (billed annually)
Get started