Silicon Valley and Wall Street are betting billions on artificial intelligence, with four major hyperscalers slated to spend about $650 billion on data centres and infrastructure this year.
Scale of the AI boom
The rapid build-out has pushed AI-linked companies to represent roughly half to 57 percent of the total value of the S&P 500, a concentration that now drives a large share of US gross domestic product growth.
Recent signs of resistance
In April, a man opposed to AI attacked OpenAI chief executive Sam Altman's home. In May, graduates entering a job market reshaped by AI met commencement speakers praising the technology, including former Google chief Eric Schmidt, with boos.
June saw Jack Clark, co-founder of Anthropic, publicly urging governments to create regulatory mechanisms to slow AI, while former president Donald Trump signed an executive order that set a voluntary 30-day national security review for frontier models.
The danger of technological determinism
Many firms assume that society will automatically adapt to new capabilities, a belief known as technological determinism. In reality, adoption depends on human incentive structures; when those are ignored, people find ways to block progress.
The Jersey Pump Principle
In the late 1940s, the automatic shut-off gas nozzle allowed drivers to fill their own tanks safely. Yet in 1949 New Jersey banned self-service fueling to protect attendant jobs, a rule that still stands today, keeping thousands of jobs alive but freezing a superior technology.
Modern parallels
Autonomous robotaxi fleets in San Francisco and Austin received regulatory approval, but the rollout ignored local drivers and emergency services. Activists discovered that placing cheap orange traffic cones on the vehicles' hoods triggered safety stalls, effectively immobilising the fleets. California regulators later suspended the licences, citing unreasonable risk to public safety.
When incentives align
History shows the opposite can accelerate adoption. When Henry Ford introduced the moving assembly line in 1913, worker turnover surged. In 1914 he doubled wages to $5 a day and reduced shift length, linking the monotony of the line to higher pay. Productivity rose and profits nearly doubled within two years.
Toyota later gave factory workers lifetime employment guarantees and installed the "Andon Cord", allowing any worker to stop the line if an inefficiency was spotted. This empowerment turned workers into partners in optimisation, leading to defect rates lower than those of many American rivals.
Implications for AI leaders
Current corporate AI deployments often aim to cut headcount, monitor output and squeeze margins. Employees therefore question why they should help the technology succeed if the only reward is higher performance pressure or job loss. Without a clear share of the productivity dividend, sabotage and slow-rolling become likely.
The automotive sector's shift to electric vehicles offers a recent example. Fears of job loss prompted the United Auto Workers to launch a 46-day "Stand-Up Strike" in September 2023. The strike produced "just transition" agreements that guarantee transfer rights, end wage tiers and bring battery-plant jobs under union contracts, removing the threat of obsolescence and smoothing the rollout.
What comes next?
For European firms and policymakers, the lesson is clear: align incentives before large-scale AI rollouts. When workers see tangible benefits, such as reduced burnout, pathways to higher-value tasks or guaranteed upskilling, they become allies rather than obstacles. Otherwise, the technology may sit idle, much like New Jersey's gas pumps, despite trillions of dollars waiting on the sidelines.