How much faith should we be putting in artificial intelligence? As large language models and generative AI have become increasingly powerful in recent years, their makers are pushing the narrative that AI is a solution to many of the world’s problems.
But Meredith Broussard says we’re not there yet, if we even get there at all. Broussard is the author ofShe coined the term “technochauvinism,” which speaks to a pro-technology bias humans often have, where we believe technological solutions are superior to anything else. In this episode, she tells New Scientist’s Sophie Bushwick that our trust in AI systems could have devastating consequences.
From discriminatory mortgage-approval algorithms, to the racial biases of facial recognition technology, to the misinformation that appears in chatbots like ChatGPT, Broussard explains why there’s no such thing as trustworthy AI. And she discusses the need for greater education about AI, to help us separate reality from marketing.