A few weeks ago, my wife and I sent the kids to their grandparents’ house and spent a weekend treating Atlanta like tourists. The experience was made all the more authentic by our hotel’s $30-per-night parking fee: Rather than shell out $60, we left the car at home.

We walked some, and we also took MARTA. But on the second night, as we prepared to leave a restaurant at 14th and Peachtree for a show at the Fox, she asked, “Why don’t we just take Uber?”

We use the ride-sharing service often when we don’t want to drive ourselves, usually for trips that aren’t easily taken by MARTA; having lived overseas and depended on public transportation, we’re comfortable taking the train. But there we were, standing a couple of blocks from one MARTA station and going within a block of another, and yet taking the train made no sense.

Our MARTA fares would be $5; the expected charge for Uber-X was $6. We’d be picked up and dropped off where we stood, rather than walking a few blocks away and taking a couple of staircases. We’d get air-conditioning, not the still air of an underground station. It was a no-brainer.

As we rode, I wondered how MARTA could adjust its business model to keep choice riders like ourselves — maybe by charging less for such short trips. But I gradually recognized the real threat ride-sharing services pose to public transportation as a “last mile” option.

For a lone person, the difference between $6 (or more) and MARTA’s $2.50 fare might be too much. But with a second person, you’re close to breaking even. With a third, you probably save money.

For longer trips, MARTA is likely safe. The cost differential is too great — although, one can easily imagine a car-pooling service that allows people to chip in with others nearby to make a longer journey more economical. In any place with enough people to justify mass transit, there are probably enough people heading in a given direction that a ride-sharing app could pool together a few of them at a time for the trip in and different combinations for the ride home, without the same handful of people having to coordinate their schedules every day.

But I digress. This is really about short trips, and the ubiquity of Uber, Lyft and their rivals makes me wonder whether a network of streetcars, like the one Atlanta envisions for itself, could ever compete with them.

The Beltline, which connects places that aren’t already well-linked by roads, still makes sense. But should we really expect a streetcar running along Peachtree to be more attractive than a car you can summon for a personalized trip? Should we spend hundreds of millions of dollars — and take lanes away from automobiles — to find out?

The ride-sharing approach would offer broader coverage and more flexibility than we will ever build out with streetcars. It is even better for low-income riders: Imagine the number of ride-sharing trips that could be subsidized for the money it would take to build, operate and maintain a decent network of streetcars.

As metro Atlanta contemplates ways to improve mobility, let’s recognize the options appearing on their own, without a dime of public money being spent. There are enough needs, and resources are sufficiently limited, that we can’t afford to duplicate any efforts.