OpenAI introduced LifeVal, an evaluation published on November 2 that assesses AI assistance with everyday responsibilities, covering areas including caregiving, education, health, shopping, personal finances, household administration, and government services. OpenAI built its task selection around the American Time Use Survey and the Multinational Time Use Study, then added a demographically representative survey of 3,995 respondents. The article notes that time spent and help wanted are different questions, and that the design challenge is to identify assistance people value rather than simply target activities that occupy the largest blocks on a calendar. It also notes that everyday tasks contain an awkward complication: different people can reasonably want different outcomes.