How Google’s AI tool joined the culture wars

Last week, Google had to pull its latest generative AI tool after it repeatedly generated ethnically diverse figures in inaccurate historical contexts, including racially diverse Nazis and Vikings. This week, the same tool provoked the Indian government’s outrage after associating Prime Minister Narendra Modi with fascism in its responses. The company is facing a right-wing backlash since launching “Gemini”: online commenters including Elon Musk have accused Google of creating “woke” AI. Building an “unbiased” model that upsets nobody is a tall order, due to political polarisation and the fact that current generative AI simply isn’t that smart – people will always find ways to “trick” such models into generating “biased” or inappropriate responses. Until AI gets smarter, tech companies will have to deal with the fact that AI is a weapon in the culture wars.