The MIT Technology Review piece opens with a telling anecdote: the CEO of an AI startup describes his own company as "self-loathing," unsure whether the technology he builds is ultimately good. That ambivalence, the author argues, is now widespread. Polls cited in the article show that more Americans expect AI to hurt them personally and society than help, with pessimism strongest among young people. In a Gallup poll, 71% of US adults opposed a new AI data center in their area—more than opposed a new nuclear plant. An NBC poll even found AI less popular than ICE.
Yet usage has never been higher. ChatGPT hit a billion monthly users in May, according to Sensor Tower, and Gemini logged 950 million in July. Pew reports that half of US adults now use a chatbot, double the share in 2023, with one in four using one daily. Across 38 OECD countries, more than a third of adults said they had used generative AI in the past three months. The numbers suggest the same people who express distrust are also the heaviest users.
The author's explanation is that the hatred is aimed not at the technology but at the companies pushing it into every corner of life, with warnings of massive social upheaval. She draws a parallel to social media, where billions kept using Facebook and Twitter despite techlash. But she sees a difference now: all 50 US states have introduced or passed AI-related bills—over 2,100 in total, a tenfold increase in three years—and open-source alternatives to the big labs could give consumers more choice and create market pressure.
The piece ends with a note of cautious hope. The Springboards CEO said there is no walking back from large language models, but they can still be made to do something different. The author echoes that, calling for clarity about what AI can and cannot do, and a technology not sold as if it is about to take over the world. With only one source, there are no conflicting accounts to weigh; the article itself acknowledges the contradiction as its central theme.