Will AI replace humans?

If you read the news, you might think robots are coming for everyone's job by Friday.

Take a breath. The real story is much more hopeful, and it has been playing out right above our heads for about sixty years.

Start with the autopilot

An autopilot holds a plane on course more steadily than any person can. Nobody argues about that. The machine really is better at that task.

And yet it never replaced the pilot.

What it did replace were jobs like the navigator, the radio operator and the flight engineer. Those jobs had something in common: look something up, work it out, pass it on. That's exactly the kind of work machines are great at.

The pilot moved up. Flying became cheaper and safer, so far more people fly today than in 1950. Every one of those flights still has a person in the left-hand seat who is responsible for it. The job didn't shrink. It got more important.

The cash machine story

Banks went through the same thing, and this time someone counted.

James Bessen wrote it up in Finance and Development, published by the International Monetary Fund in March 2015. When cash machines arrived, each city branch needed fewer tellers. Between 1988 and 2004, the average city branch went from needing 20 tellers to 13.

You'd expect teller jobs to disappear. They didn't. Branches became so much cheaper to run that banks opened 43% more of them in cities, and the total number of teller jobs held up.

The tellers who stayed weren't counting cash faster. They were opening accounts, solving problems, giving advice and remembering customers' names. David Autor, an economist at MIT, explained why in the Journal of Economic Perspectives in 2015. When a machine makes some steps of a job cheap, the value moves to the steps the machine can't do.

"A bank teller who can tally currency but cannot provide 'relationship banking' is unlikely to fare well at a modern bank."

David Autor, Journal of Economic Perspectives, 2015

Counting cash went to the machine. Knowing the customer became the job.

A simple test for your own job

Three economists, David Autor, Frank Levy and Richard Murnane, wrote down the rule behind all of this in the Quarterly Journal of Economics in November 2003. It still works today.

Ask one question about any task: could you write down the rules for it in advance?

If yes, a machine will probably do it, and sooner than you think. If no, it stays with a person.

Try it on your own week:

• Searching for the right person to work with? Rules. A machine can do it.
• Chasing an unpaid invoice? Rules. A machine can do it.
• Writing a first draft? Mostly rules. A machine can help a lot.
• Telling someone kindly that their idea won't work, and offering a better one? No rules at all. That one is yours.

Why this is good news

Here's the best part. The rule-following parts of most jobs are the parts nobody enjoys.

Nobody chose their career because they loved filling in forms or chasing payments. When those tasks go, your week isn't emptier. It's full of the work you were actually hired for, and probably the work you like most.

There's a second piece of good news hiding in both stories. When something useful gets cheaper, people want more of it. Cheaper flights meant more flights. Cheaper branches meant more branches. And more of everything means more need for people to make the calls and take care of the customers.

What to do about it

The move is simple, and it's the same one the pilot made.

1. Hand over the checklist work. Let machines do the searching, the drafting, the chasing and the reporting.
2. Spend the time you save on judgement. Deciding, persuading, building relationships and owning the result.
3. Get good at directing the tools. Knowing what to ask for, and when the answer isn't good enough, is a skill worth more every year.

Do that, and AI stops being a threat. It becomes the best assistant you've ever had.

Where Lumienzo fits

We built Lumienzo around exactly this idea. Lumi runs the rule-following side of influencer marketing on its own. It learns about your brand and product and researches what's trending. It writes the campaign plan and finds creators using data from accounts they connect themselves. Then it tracks every post and reads every comment. You just review and approve.

What it doesn't do is make the video, earn an audience's trust or put its name on the work. Those stay with people, which is the whole point. See how it works for creators and brands, or read about what AI actually replaces in influencer marketing.