Automation is eroding the skills of drivers and mathematicians

Automation is reducing how much drivers and mathematicians understand about the systems they rely on, ERR commentator Kristjan Port argues in a technical commentary for R2 Portal. He likens the loss of control to a Faustian bargain, in which absolute power and limitless knowledge come at the cost of control over one’s soul.

Earlier generations of drivers needed not only driving skills but also knowledge of how cars were built, maintained and repaired. Getting from A to B depended on the car starting, having enough oil and keeping its tyres properly inflated. A faulty carburettor or a lost spark from the ignition coil could leave a car stranded, so drivers carried tools and knew how to use them.

Fuel-injected engines and computer control made cars considerably more reliable, while fewer drivers learned how the vehicles worked. Mechanics and repair shops increasingly relied on computer diagnostics. In some of the newest cars, drivers can no longer look under the bonnet; they may only have access to small openings for adding windscreen washer fluid and fuel, whether liquid or electric. Computers monitor the car’s condition over the internet.

Port sees a similar shift in mathematics. Mathematicians once worked with paper and pencil, handling abstract concepts and making connections. They calculated intermediate results by hand, checked edge cases and had to remember a great deal. Early computers automated some of these tasks, but mathematicians still formulated conjectures and strategies before automated proof systems checked them.

Mathematicians’ work became more valuable because finding clever solutions still required human intelligence. Now artificial intelligence can produce checked formal proofs, some hundreds of pages long, in a short time. Some proofs use strategies that are new to mathematicians, who may need months to understand and verify them.

Port says early car driving and mathematical work shared a process in which people developed expertise by overcoming failures and solving problems. Obstacles were not just challenges but also helped shape problem-solvers. Finding a way out of traffic jams built skills and intuition, while diagnostic computers and AI-generated proofs reduce the personal effort needed to become an expert.