What happens when artificial intelligence learns from a world already shaped by inequality?
In this episode of The Revolution Resolution, Dr. Anna Malaika Tubbs examines how AI can absorb and reproduce the biases, stereotypes, and hierarchies embedded in the data used to train it.
Beginning with the viral Lensa AI avatar phenomenon, this episode explores how technology can reinforce existing ideas about gender, race, beauty, power, and desirability — from hypersexualized images of women to racial bias and narrow representations of who gets imagined as a leader.
But the implications extend far beyond AI-generated images.
As artificial intelligence increasingly influences hiring, search, surveillance, recommendation systems, content moderation, and everyday decision-making, we have to ask a much bigger question:
What kind of world are we training technology to reproduce?
Because AI does not emerge outside society. It learns from us.
And if we don't interrupt the hierarchy embedded in our existing systems, we risk automating it into the future.
If this episode resonates, share it with someone thinking critically about technology, representation, and the future we're building together.
