Forget trying to pick an AI-proof career for your kids. Instead, make sure they can still do math with a pencil and paper and introduce them to artificial intelligence early on.
That’s the advice of British economist Daniel Susskind, a senior research associate at Oxford University’s Institute for Ethics in AI, a digital fellow at the Stanford Digital Economy Lab and the author of “A World Without Work.”
In his new book, “What Should My Children Do? A Human’s Guide to the Age of AI” (Penguin Press, out now), the father-of-three offers some surprising advice for parents worried about their children’s professional futures. In an exclusive interview with The Post, he also came down hard on New York City Mayor Zohran Mamdani’s one-year ban on student-facing generative AI in public schools through eighth grade, announced in September.
“I completely understand the instinct behind the decision,” Susskind said. “But I think it is a big mistake.
“It is letting down the next generation if we do not prepare them for the world in which they will actually live, coddling them in an artificially tech-free environment that doesn’t reflect their future,” he added. “I do hope the ban does not extend beyond a year.”
When it comes to what parents should be doing to prepare their children for adulthood, Susskind has three recommendations.
First off, he said, parents should stop looking for a career that AI can’t touch, as we just don’t know what that will be.
“The advice that makes me wince is anything that draws on the idea of ‘future-proofing,’” Susskind said. “We just can’t predict what exactly the world of work will look like in the decades to come and so what advanced skills will be most valuable for kids to learn today.”
Recall that, not so long ago, it seemed that learning to code was the answer to ensuring a bright future for your child. In 2014, England introduced a new computing curriculum that included teaching children as young as 5 to code. In 2019, Joe Biden suggested that even coal miners should learn computer programming.
Then came generative AI, which turned out to be especially good at the very skills we’d been telling our kids they needed to learn. By April 2025, Microsoft CEO Satya Nadella was reporting that AI was writing as much as 30% of the code for some of the company’s projects.
So should parents still fork over hundreds of thousands of dollars to put their kids through medical or law school?
“It might sound surprising, but my advice to that 15-year-old is still to take the traditional route,” Susskind said.
Some of the most promising applications of AI, he explains, come from pairing people who understand the technology with people who’ve spent years studying the problems it could help solve. A medical education gives an aspiring doctor the expertise to recognize which clinical problems are worth tackling and whether a machine has come up with a medically sound solution.
“But then, as you move through your career, be open-minded about what your ‘job’ looks like, or what ‘organization’ employs you,” he said.
His second recommendation might sound familiar to anyone who went to elementary school in the 1950s: Get back to basics.
“However the future turns out, and whichever advanced skills do turn out to matter, these basic skills are going to be vital as they are the building blocks for everything else,” he said.
Children will need to know enough math to recognize when a machine has made a mathematical error. They’ll need to read critically enough to catch a chatbot inventing facts. The convenience of having a machine write a persuasive argument for them could come at the expense of ever learning how to construct one themselves.
Susskind believes the solution can be found in a nearly half-century-old British government report about calculators.
In 1982, mathematician Wilfred Halliday Cockcroft proposed that students should be tested on their ability to do math both with and without calculators. Calculator use was controversial enough in the US that, four years later, a group of math teachers took to the streets of Washington, DC, chanting, “Cal-cu-la-tors later, we shall not be mov-ed.”
Susskind would like to see a similar approach adopted for AI across every school subject. An English teacher, for example, could have students interpret a poem on their own and then use AI to tackle a more challenging piece of literature.
He follows much the same policy at home with his children, “embracing technology in the right moments but also religiously protecting time without any of it.”
Susskind supports keeping smartphones away from children until they’re 14 and social media until they’re 16.
But, his third recommendation is to introduce children to AI while they’re still in elementary school.
“You can be very concerned about the effects of the former,” he said of social media, “but also try to make sure that those worries don’t morph into a general technological conservatism and knee-jerk bans on AI.”
When his 6-year-old daughter was getting bored learning her multiplication tables, Susskind asked her to invent a computer game that would make practicing more fun. She imagined a giant rainbow in the sky, with nuggets of gold delivered by flying unicorns every time she got an answer right.
Susskind, with little experience in computer programming, asked ChatGPT to build it. After a little troubleshooting, his daughter was playing the game she’d dreamed up just minutes earlier.
“For me, this was an astounding moment,” he recalls. “What seemed remarkable to me felt commonplace to her. And why would it not? This powerful technology is just another part of the furniture of her childhood.”
More recently, he’s been using AI to create books that help his 5-year-old son learn to read, with stories tailored to the boy’s latest obsessions: Lego and a British warship, the HMS Belfast.
His youngest child, who’s 2, has yet to benefit from her dad’s AI experiments.
“My focus there is getting her to sleep through the night!” he said.
Read the full article here














