AI Economy Quiz: How Ready Are You for the Transformation Ahead?

Why are skilled trades like electricians and plumbers considered AI-resilient occupations?

Skilled trades are fully automated but workers are still needed to supervise the machines Skilled trades require no judgment and are therefore not worth automating AI models already work as plumbers and electricians in smart homes They require physical dexterity in variable, unpredictable physical environments — something current AI cannot perform — and industry aging plus declining new entrants is creating wage-driving shortages

A restaurant receives 20 new Google Reviews per week. Responding to all of them takes too much time. What is the best AI-assisted approach?

Use AI to respond automatically to all reviews without any human review Consolidate all responses into one weekly post Use AI to generate personalized response drafts for each review that the owner quickly reviews and publishes Respond only to negative reviews — positive reviews don't need responses

What is the difference between 'high AI exposure' and 'high substitution risk' for a role?

Substitution risk refers only to blue-collar roles; white-collar roles only face exposure High exposure means AI significantly affects the tasks in a role; it may complement the worker (boosting productivity) or substitute for them. Risk depends on whether AI augments or replaces the majority of tasks High exposure and high substitution risk mean the same thing High exposure always means high risk — any role AI touches is at risk of disappearing

Which statement about AI API output tokens is correct?

Output tokens are always cheaper because the model is just retrieving stored answers Input and output tokens are always priced identically at major providers Output tokens are typically more expensive per token than input tokens Output tokens are not counted toward API costs at the Sonnet/GPT-4o tier

A team builds a production application on an AI API with a stated 99.5% monthly uptime SLA. Approximately how many hours of potential downtime should they plan for per year?

~44 hours per year 0 hours — 99.5% means near-perfect reliability ~5 hours per year ~0.5 hours per year