AI and Work

Your Own Experience

On Tuesday you used AI to analyze data, create visualizations, and present findings. Think about that experience, and any other times you've used AI for work or school.

Discuss with your partner: How did using AI for data analysis feel? Did it make you more capable, or did you feel like you were outsourcing your thinking? Have you used AI for work in other classes or jobs? When AI does something well, does that feel empowering or threatening?

What's Actually Happening to Work

Thompson's coders say they're happier and more productive with AI. They spend less time on tedious boilerplate and more on interesting problems. But Jack Dorsey fired half of Block's employees because AI tools paired with smaller teams enabled "a new way of working."

Lowrey offers a framework: are you a horse (permanently displaced by a better technology) or coal (temporarily disrupted but ultimately in greater demand because efficiency drives new uses)?

Discuss with your partner: Can both things be true: workers happier AND expendable? Who in Thompson's piece is a "horse" and who is "coal"? What determines which?

Horse or Coal?

Lowrey's framework: horses were replaced by tractors and never adapted. Coal was supposed to be replaced by efficiency, but the Jevons paradox meant demand actually increased. Radiologists were predicted to be obsolete by 2021, but they survived because improved imaging created new demand AND FDA regulation slowed AI deployment.

Discuss with your partner: Apply the horse-vs-coal test to your own future career or field of study. Are you more likely to be enhanced by AI or replaced by it? What about people who don't go to college? Lowrey says factors beyond skill matter: regulation, professional licensing, consumer preferences. What structural factors, beyond just being good at your job, might protect or endanger your career?

Who Decides?

Cassidy describes two uses of AI: one that helps electricians diagnose problems faster (pro-worker) and another that surveils Amazon warehouse employees' every movement (anti-worker). Same technology, opposite effects. The difference isn't the AI. It's who decided how to deploy it and in whose interest.

Thompson's Block employees had no say in Dorsey's decision to fire them. Lowrey notes that radiologists survived partly because FDA regulation slowed AI deployment in medicine.

Discuss with your partner: Who should decide how AI is deployed in workplaces? The companies building it? The workers affected? Government regulators? What would "pro-worker AI" look like at Illinois, in your major, or in a campus job?