Anthropic interviews 80,000 people in depth: where did AI save you the time

Anthropic conducted an in-depth one-to-one interview of 80,508 people in 159 countries, and found that what was most desired was not a stronger AI but more time, with one third of the respondents' core desire to be able to spend time with their families; the report found that the benefits and harms of AI were simultaneously occurring with the same person: people who benefited from learning help were at the highest risk of being aware of degradation, those who spent time in AI were being speeded up, and those who supported them were three times more likely to be relying on AI than ordinary people; 16.31 TTP3T people voluntarily acknowledged that their ability to think was declining, and 191 TP3T believed that AI was not fulfilling its commitments; the report revealed that AI benefits were immediately known and slow to harm, and that when cognitive outsourcing became the default option, it was humans who were most likely to surrender their will to independently judge。