Science Editor, The New York Times; Author, The Secret Life of the Grown-up Brain; The Primal Teen
Machines That Work Until They Don't

When I'm driving into the middle of nowhere and doing everything that the map app on my smartphone tells me to do without a thought—and I get where I am supposed to go—I am thrilled about machines that think. Thank goodness. Hear Hear.

Then, of course, there are those moments when, while driving into the middle of nowhere, my phone tells me, with considerable urgency, to "Make a U-turn, make a u-turn!'' —at a moment on the Grand Central Parkway where such a move would be suicidal. Then I begin think that my brain is better than a map algorithm and can tell that such a U-turn would be disastrous. I laugh at that often life-saving machine and feel human-like smugness.

So I guess I am a bit divided. I worry that, by relying on my map app, I am letting my own brain go feeble. Will I still be able to read a map?? Does it matter?

As a science editor and daughter of a mechanical engineer, who trusted machines more than people, I would think I would automatically be on the side of machines. But while that mechanical engineer was very good at figuring out how to help get Apollo to the moon, we also had a house full of machines that worked, sorta. A handmade stereo that was so delicate you had to wear gloves to put a record on to escape the prospect of dreaded dust, etc. We are all now surrounded by machines that work, sorta. Machines that work until they don't.

I get the idea of a driverless car. But I covered the disaster of Challenger. I think of those ill-advised U-turns. I don't know.

On the one hand, I hope the revolution continues. We need smart machines to load the dishwasher, clean the refrigerator, wrap the presents, feed the dog. Bring it on, I say.

But can we really ever hope to have a machine that will be capable of having—as I just had—five difficult conversations with five other work colleague human beings? Human beings who are lovely but have, understandably, their own views on how things should be?

Will we ever have a machine that can get a 20s something do something you think they should do but they don't? Will we have a machine that can, deeply comfort another at a time of extreme horribleness?

So, despite my eagerness for the revolution to continue, despite my sense that machines can do much better than humans at all sorts of things, I think, as an English major, that until a machine can write a poem that makes me cry, I'm still on the side of humans.

Until, of course, I need a recipe really fast.