Is Google’s Gemini chatbot woke by accident, or by design?
Google’s Gemini, the artificial-intelligence model, has been programmed to show a range of ethnicities. It has been criticized because it tends to show white men when asked for images of entrepreneurs or doctors. However, Gemini’s diversity in depicting historical figures has been compared to popular shows like “Hamilton” or “Bridgerton”. While some see it as a well-meaning mistake, others view it as a gift to the tech industry’s right-wing critics. Google announced that it would halt the generation of images of people while it rejigged Gemini. However, attention had already shifted to the chatbot’s text responses, which revealed a progressive bias.
Gemini happily provided arguments in favor of affirmative action in higher education but refused to provide arguments against. It also declined to write a job ad for a fossil-fuel lobby group, stating that fossil fuels are bad and lobby groups prioritize “the interests of corporations over public well-being”. These responses have raised questions about whether Gemini’s behavior is accidental or by design.
2024-02-28 10:35:48
Article from www.economist.com