When Intelligence Becomes Cheap, Judgement Becomes Valuable
For the first phase of the generative AI revolution, companies rewarded adoption. Employees were encouraged to experiment, automate, prompt, accelerate and produce. Now a more difficult question is emerging. Once almost everyone has access to increasingly capable artificial intelligence, what actually distinguishes one worker, manager or organisation from another?. Ernst & Young’s US business has supplied an unusually concrete answer. In August 2026, EY announced a $100 million investment in employee rewards aimed not merely at technology adoption but at business acumen, judgement, adaptability, innovation, future-focused skills and exceptional client impact. The announcement is significant because compensation systems reveal what institutions believe has value. EY is effectively putting money behind a proposition that many organisations are only beginning to understand. The AI economy will require technical capability, but technical capability alone will not be enough. This is not the first time a machine has changed the value of human labour. Steam power changed physical production. Electricity reorganised factories. Mechanisation transformed textiles. Computers reorganised information work. ATMs automated transactions. Smartphones subsequently reorganised banking itself. Each revolution eliminated some activities, created others and, most importantly, changed what remained valuable for people to do. AI belongs to this history, but it introduces a profound difference. It reaches directly into work once assumed to demonstrate intelligence itself. The resulting challenge is not human versus machine. It is far more demanding. Organisations must decide what machines should do, what humans must continue to understand, how people will develop judgement when machines increasingly perform the tasks through which judgement was once learned and how value will be distributed when technology radically increases the productive capacity of labour. The machine can increasingly produce an answer. Someone still has to decide whether the answer deserves to become reality.
