The Demonstration Works.

What Happens When Everybody Buys One?

I watch a humanoid robot carry a box across a room. It puts the box down without dropping it or falling over.

That is an accomplishment. Engineers have spent years getting machines to do things most of us learned before kindergarten, and I don’t mean that as an insult. Walking, seeing, gripping, and adjusting to unexpected obstacles are remarkably complicated activities.

But while I’m watching, my mind usually wanders beyond the demonstration.

What does the robot cost? How much work can it do in a day? Who helps when it gets confused? And if I owned a small business, would buying one leave me better off?

Then comes the question that interests me most: What happens when these machines stop being unusual?

I Have Always Had Trouble Staying Inside the Forecast

During my working years, one of my strengths was making an observation and quickly exploring what it might make possible. Sometimes those possibilities went considerably further than the original observation appeared to justify.

That was also why it helped to have colleagues willing to challenge me.

I used to chide my financial forecasting friends about the slope of their curves. They were generally comfortable projecting growth of three or four percent. A particularly adventurous forecast might approach ten percent.

Technology forecasting could take us into different territory. A combination of technical advances, changing costs, social behavior, and new business arrangements could suggest growth measured in multiples rather than a few percentage points.

The difficult part was figuring out what would have to happen for those possibilities to become real.

In 1998, our discussions about the internet explored changes that seemed extravagant at the time. Then came the acceleration, followed by the crash in 2000 and 2001.

It would have been easy to conclude that the extravagant forecasts had been disproved.

But when we looked back at our forecasts in 2005, we found that in several areas we had underestimated the impact.

The crash and the transformation were both real. They were simply telling us different things.

One Drone Is Interesting. Thousands Raise Different Questions.

That experience still influences what catches my attention.

In our recent drone articles, I kept coming back to a question that I don’t think gets enough attention in the public discussion.

Suppose a dozen organizations each decide to operate a few thousand drones.

Some carry meals. Others deliver medicine, inspect power lines, survey property, or move supplies. They share the sky with aircraft already serving personal, business, commercial, and emergency purposes.

The question expands from whether a drone can complete a flight to how all those flights fit together.

Who resolves conflicting routes? What happens when weather changes or communications fail? How does an emergency helicopter get priority? Who is responsible when separately operated systems make decisions that interfere with one another?

I am not suggesting that nobody is working on these problems. What I want to understand is how the pieces will fit together when operations become commonplace.

A successful delivery demonstrates that something can work. Repeating it thousands of times, among other people doing the same thing, creates a different challenge.

The Robot Has to Earn Its Keep

Robotics raises a related set of questions.

The humanoid demonstrations are fascinating. But a business owner needs to know more than whether a machine can perform a task.

Can it do enough useful work to justify its purchase, maintenance, software, supervision, and occasional rescue? Will it help existing employees accomplish more? Does the workplace need to be rearranged around it?

And does the machine need to look like a person?

Sometimes that shape may be useful in spaces built for human bodies. In other situations, a specialized machine may be a better answer. I would be quite happy for a gardening robot to look nothing like me, provided it could distinguish a weed from something I had paid to plant.

Then there is the matter of working together. A robot that performs well by itself still has to fit into a business with people, equipment, schedules, and perhaps other robots from different suppliers.

Someone has to make the whole operation work. That person may discover that the demonstration left out a few expenses or processes.

This Is Where My Optimism Starts Asking Questions

I remain interested in what these technologies could make possible. Machines that take over difficult tasks, help people remain independent, or allow a small business to offer something previously impractical could create substantial value.

But I want to explore the steps between a promising invention and that useful result.

Who can afford it? Who benefits? What new work becomes possible? What supporting services will someone have to invent?

Those were productive questions during the internet years. I suspect they remain productive now.

I no longer have a research organization and a roomful of colleagues to help investigate every possibility. These days, I have my experience, my curiosity, some assistance from artificial intelligence, and readers who may see things I have missed.

That last group could be particularly useful.

If you operate a business, manage equipment, or simply have a household chore you would gladly surrender, I would like to hear your answer:

What would a robot have to do reliably before you would pay for one—and what would you be willing to pay?