We see people, computer-seep pixels, like they don't see what we see.

It's note to self, I'm a new summer ody, and amazing things are happening to the way

we live.

We eat, we get around.

The salad you ate for dinner last night?

The lettuce may have been screened by a lettuce bot, mounted on a tractor that can see whether

the plant needed spraying or not.

Can't get to the dermatologist?

No problem.

Pretty soon an app will be able to check that mole on your arm.

Heck, if you live in Pittsburgh or Singapore, you already might be handling a self-driving taxi.

Maybe write this minute.

But for you to believe an app that claims your mole is pre-cancerous, or trust that your

self-driving car is taking you home the long way for a good reason, you need to understand,

at least a little bit, how it made that decision, right?

Just knowing the basics behind buzzwords like artificial intelligence and neural networks

and deep learning can help us understand why these algorithms sometimes make mistakes.

And then we can help make them better, even if we aren't computer engineers.

And that's what today's episode is about.

A nasty, racist mistake that an algorithm made.

But really, the algorithm can't be blamed.

It's much bigger than that.

...