Beyond Today

Could AI do your job?

·23 min·2 clips
A proud hansom carriage driver confronts AI competition while tech leaders claim innovation means 'we are flying today'.
The episode opens with host Tina Daheley posing the question 'Could AI do your job?', contrasting utopian tech visions with a hansom carriage driver's fear of displacement. Guest Daniel Suskind, an Oxford economist and author of 'A World Without Work', discusses why people overestimate their job skills, citing an anecdote where a tax accountant claimed clients valued personal touch, but Suskind countered they primarily want affordable, timely service. Suskind explains AI's capabilities, such as a Stanford system that diagnoses skin cancer from freckle photos using pattern recognition on 130,000 cases, performing tasks in fundamentally non-human ways. The conversation addresses historical economic shifts, noting that while new industries emerge, they create fewer jobs relative to the past, with only 0.5% of US employment in 2010 in industries created that decade. Suskind critiques John Maynard Keynes' 1930 essay predicting a 15-hour workweek, arguing Keynes overlooked the distribution problem where prosperity concentrates rather than spreading evenly, linking automation concerns to rising inequality. The episode explores whether 'soft skills' like empathy protect jobs, mentioning affective computing and examples like a Chinese professor using facial recognition to detect student boredom or systems identifying courtroom lies. Suskind describes 'frictional technological unemployment' involving skills, place, and identity mismatches, such as white men avoiding pink-collar jobs like nursing due to perceived identity conflicts. The paradox of socially valuable jobs like caring professions being low-paid is discussed, alongside the relationship between work and meaning, referencing ancient philosophers who saw work as grubby and hunter-gatherers working fewer hours. Suskind advises preparing through education focused on skills to compete with or build machines, criticizing current policies for teaching tasks AI excels at. The episode concludes with political implications, suggesting regulating tech companies requires moral philosophy beyond economics, as their free services wield power without transparency.
Listen to the show on