The Technology Trap

From a great historical distance, industrialisation looks like a success story: higher productivity, new goods, increasing prosperity. For many people whose work was devalued, wages depressed or communities shaken, the same development initially looked different. Carl Benedikt Frey uses this tension to place today's automation debate in context.

His core message: a technology can create long-term societal benefit while still distributing substantial costs unequally during the transition. Looking only at later average income fails to explain resistance, political polarisation and real losses.

1. Labour-replacing and labour-complementing technology

Frey distinguishes analytically between technologies that complement human abilities and those that automate tasks and can displace workers. In reality, many applications lie between these poles. An AI system can make professionals more productive, take over individual routines and simultaneously create new control or qualification requirements.

The question ‘How many jobs will disappear?’ is therefore too broad. Better questions are: which tasks are changing? Who gains scope for action? Who loses income, bargaining power or access? What new activities emerge, and who can access them?

2. Technology works through institutions

Machines do not determine their societal consequences alone. Ownership, education, labour law, competition, regional infrastructure and political decisions influence how gains are distributed and transitions shaped. Two countries or companies can deploy the same technology very differently.

This perspective strengthens systems thinking. A training programme can be sensible yet remain ineffective if new jobs emerge far away, caring support is lacking or selection processes devalue previous career histories. A good analysis considers technology, people and institutions together.

3. Resistance contains information

Historical protests against machines are often portrayed as irrational hostility to technology. Frey's account reveals that affected people expected real and sometimes lasting losses. This does not mean preventing every innovation. It means treating objections as signals of distributional issues, lack of voice and uncertain transitions.

Transformation skills therefore also mean recognising conflicts early and involving affected people in design. People are more likely to accept change when they have influence, a realistic outlook and protection from one-sided risks.

Your three-minute exercise: winners, losers, design

Take an automation proposal being discussed in your environment:

1. Minute 1 – tasks. Which specific activity is being complemented, changed or replaced? Avoid broad occupational labels.
2. Minute 2 – distribution. Who gains time, income or power – and who bears the learning effort, uncertainty or loss of control?
3. Minute 3 – transition. What participation, training or protection would make the change fairer and more productive?

What we take from this

The book adds a societal perspective to Future Skills. Willingness to learn matters, but must not shift structural responsibility onto individuals. Within WEITER7 the reality check therefore includes: what conditions support development – and which barriers must organisations or policymakers change?

Which costs of technological change remain invisible when we look only at efficiency?

Context: This insight distils Frey's historical argument in our own words. Historical comparisons explain patterns but do not automatically transfer to today's AI. Whether technology mainly complements or replaces work is not solely a property of the tool, but depends on its context of use. The exercise, Future Skills links and WEITER7 application are our own interpretations.

Sources and further reading: Carl Benedikt Frey's official website ↗ · Oxford Martin School: The Technology Trap ↗ · International Labour Organization: Book lecture ↗