AI could hollow out the next generation of workers

• Tom Slater said AI could hollow out the next generation of skilled workers by eroding their skills.
• The investor said workers still need struggle and repetition to build expertise.
• Slater warned short-term AI productivity gains may create long-term skill erosion.

AI may make young workers faster and more productive, while also quietly stopping them from becoming experts at all.

That’s the warning from Tom Slater, the manager of Scottish Mortgage Investment Trust and a partner at Baillie Gifford, who said that companies embracing AI too aggressively risk hollowing out the pipeline that trains future professionals.

Speaking on the “Merryn Talks Money” podcast on Monday, Slater said that unless we invest in the next generation and ensure that they learn foundational and analytical skills, businesses could end up with workers who know how to use AI tools but lack the foundational expertise needed to judge whether the output is actually correct.

His concern echoes that of AI researchers and some tech leaders who have warned that overreliance on AI could gradually erode the very skills workers need to perform and think critically on the job.

Innovation theorist John Nosta said AI is training humans to think backward by giving polished answers before people fully understand them. AI researcher Vivienne Ming said that most users are relying on AI to think for them rather than using it to deepen their reasoning.

Slater said that younger employees need to struggle through difficult tasks at work to develop expertise.

“You would just get better at using AI, but you will confuse that with thinking that you’re good at the actual topic,” he said.

Slater pointed to research suggesting that overreliance on AI can weaken learning.

He cited an MIT study from last June that tracked 54 participants over four months and found that people who used ChatGPT to write essays showed lower brain engagement and often struggled to remember or quote from work they had produced only minutes earlier.

The problem, he said, is that AI improves short-term productivity while quietly weakening long-term capability.

That dynamic could create a dangerous incentive for employers, he suggested. Companies can save money by reducing entry-level hiring and leaning more heavily on AI systems, but Slater believes that strategy amounts to a “false economy.”

“It seems to me very shortsighted,” he said of allowing the current generation of experienced professionals to use these tools to get a “productivity bump” without thinking about the next generation.

Slater compared the issue to pilots learning to fly manually before relying on autopilot systems. Professionals, he said, still need to develop deep expertise before AI can genuinely enhance their work.

“The people who will thrive are not those who use AI the most,” Slater said, “but those who can still think without it.”