Elon Musk agrees AI is learning the one thing humans thought it couldn't do

Elon Musk believes artificial intelligence could eventually do almost every job better than humans, leaving people free to work only if they want to. Now, the billionaire has agreed with a claim that challenges another perceived limit of AI, its ability to develop human taste and creative judgement.Benji Taylor, Head of Design at X and SpaceXAI, just wrote on X, “I no longer see any compelling reason to believe that taste is some permanent boundary LLMs won’t cross. It’s already happening.” Musk responded with a one-word reply, saying "True."Taylor's post raises a question that goes beyond design. If AI can learn to judge what looks good, create compelling products and make decisions that once required human experience, how much of an advantage will people retain as machines become more capable? Elon Musk openly agrees with the view that AI will take over human tasks. Elon Musk thinks work could become optionalMusk has repeatedly warned that AI and robotics could eventually make human labour unnecessary for much of the economy. In a July 2026 interview with The Economist, he said AI would be able to perform any job better than a person, adding that this would happen particularly quickly in digital work, including tasks performed on computers and phones. However, Musk's vision of the future is not necessarily one in which everyone remains unemployed and struggles to survive. He believes advances in AI and robotics could create an age of abundance, in which people have access to goods and services without needing to work for a living.At the US-Saudi Investment Forum in January 2026, Musk said work would become optional, comparing it to playing sports or video games. He used gardening to explain his point that people can buy vegetables from a shop, but they can also grow their own if they enjoy doing so. The home-grown vegetables may not be as perfect as those available in stores, but they can still offer personal satisfaction. In other words, Musk imagines a future in which people might choose to work because they enjoy it, rather than because they need the money to survive.Musk has also proposed universal high income, going beyond universal basic income (UBI). In September 2026, he said AI and robotics could make abundant goods, services and better healthcare available to everyone. This vision, however, depends on how the wealth generated by automation is distributed.Sridhar Vembu has a different idea for the AI eraZoho founder Sridhar Vembu has also questioned what humans will do if AI takes over intellectual work, including software development and mathematics. But he disagrees with the idea that universal basic income is the answer.In an October 12 post on X, Vembu called UBI an “unimaginative and dystopian vision." Instead, he suggested that people could earn a living through activities involving nature, nurture, culture and scripture. These include restoring soil and water, raising children, caring for elderly people, making music, playing sports and practising spiritual traditions.Vembu's argument is that these activities could become meaningful sources of livelihood if AI makes the production of goods and services far cheaper and society distributes the resulting wealth more widely. His proposal differs from Musk's emphasis on universal high income as Vembu puts greater weight on finding valuable work beyond conventional intellectual jobs.Anthropic research shows not every job is equally exposed to AIThere is still a difference between AI being capable of performing a task and replacing an entire occupation. Research published by Anthropic on September 30 found that large language models and robots could potentially perform tasks accounting for around 80 percent of working time in the US economy. But current robots remain expensive and have limitations in physical settings.The research found that nursing and general repair work were less exposed to current robots than driving and warehouse jobs. Work involving human interaction, physical contact and fine motor skills remains difficult for robots in many settings.This does not mean such jobs are permanently safe from automation. It shows that the impact will depend on how quickly AI and robotics improve, how much they cost and whether people want humans involved in certain activities.- Ends