In the long run,

the facts are on the side of the optimists.

To paraphrase Martin Luther King Jr., the arc of history bends toward progress. But progress doesn't just happen. People work hard to discover ways forward.

The What Works Initiative
highlights positive outcomes on difficult issues – and how people achieved them.

A Progress Postcard:

The threats posed by artificial intelligence range from the local – via the spread of power-hungry data centers – to the existential – as industry insiders warn that AI agents could wipe out humanity within a decade.

Then there’s the jobs.

AI’s very promise of a productivity supercharge across the whole economy raises the specter of massive unemployment. The Luddites, after all, were right. The mechanical looms they smashed two centuries ago went on to take away their livelihoods as textile makers and thrust many of them into deep poverty. In the long run, England and its people were richer for the automation, but it was a long, painful run.

The good news is that with AI, there is still time to do it better. In spite of the gigabit-fast adoption of AI, the job apocalypse has yet to appear. In fact, The Economist recently estimated that AI has created about a million jobs in the US since the tech began taking hold in mid-2023. It put the losses at about 200,000.

The new jobs pay well too. The Economist noted that installation and maintenance jobs at data centers advertise pay 40 percent higher than comparable jobs elsewhere.

Employment rates for young college grads, for the overall 20-to-24 age population, and for the 25-to-54 population are holding at about the same levels as before ChatGPT ever hit the streets.

At least one study shows that businesses that have adopted AI hire more people than peer companies that haven’t.

Another study found something counterintuitive about who benefits from AI adoption. A study of AI use among patent attorneys found that senior attorneys using AI became much more productive. Junior attorneys without long experience actually lost ground. In other words, AI is great leverage for those who already know their stuff – who know what to ask and can evaluate the responses. If you don’t, using AI won’t teach you.

Sources: The Economist; BLS; NBER Working Paper 35720

Our starting issue: