What can psychology teach us about AI’s bias and misinformation problem?

Knowledge may be power. But what if the information that leads to that knowledge is wrong?

To Celeste Kidd, assistant professor of psychology at UC Berkeley, the answer is simple: It’s dangerous and perhaps the most concerning aspect of generative AI’s rapid expansion. 

Systems like ChatGPT, Stable Diffusion and DALL-E have rippled across the planet. Millions of people have used these platforms for amusement, incorporated them into their professional workflows and turned to them for quick answers on subjects both trivial and complex.