Don't Bogart My Blog: Sunday Best: This algorithm will tell you what to be for Halloween

This is the scary side of machine learning.

View in Browser | Add nytdirect@nytimes.com to your address book.

Sunday, October 28, 2018

NYTimes.com/Opinion »

The air is crisp, Jack-o'-lanterns are omnipresent, and soon children will dress up for an evening in exchange for candy. I'm certainly ready for Halloween, the one time of year I claim to not fear horror films and binge-watch them while actually scrolling through Instagram (where, as of last week, you can now find the compelling art and strong perspectives of The Times's Opinion department at @nytopinion).

But sometimes, it's not ghosts or ghouls that we find the most terrifying. Sometimes, it's our realities that haunt us: our lives, the systems that surround us, even our own bodies. Happy Halloween. — Alexandra March

Your Next Halloween Costume? The Whimasy Arat

By JANELLE SHANE

Or perhaps sexy marijuana bee, Ruth Bader Hat Guy or frog-Shadown fiking Cesterian are more appealing. These are some of the characters that a machine learning algorithm generated when we gave it 7,182 costumes and asked it to come up with Halloween costume ideas.

I Believe in Ghosts

Laura Lannes

By KAITLYN GREENIDGE

"For now, we are learning to live with our ghost. Kirsten says the ghost works beside her when she's home in the afternoons. She can hear the ghost opening and closing the children's dresser drawers. Sometimes, she tells me, the ghost folds her laundry."

The Identity Crisis That She Never Saw Coming

Mathilde Aubier

By JENNIFER LATSON

Jennifer Latson has been held captive by debilitating migraines for over 20 years, more than a third of her life. They have dictated her plans and decided her career. Finally, there may be a cure, but it doesn't come without apprehension: Who do you become when one of your defining characteristics vanishes?

There Is No Shame in Having a C-section

A baby born via C-section. Sascha Rheker/VISUM, via Redux

By HONOR JONES

"Patients should have better access to data on C-sections, so that they can avoid doctors and hospitals that order them disproportionately. And we should all demand that doctors worry as much as mothers about whether surgery is justified."

How to Quit, Come Back, Then Beat the Odds

Allie Kieffer, training in Buffalo, placed fifth in the New York City Marathon last year. It was an astonishing result for an unsponsored 30-year-old who had never had a top finish in a major race. Libby March for The New York Times

By LINDSAY CROUSE

Last year, Allie Kieffer was a 30-year-old hobby runner when she shocked everyone by finishing fifth in the New York City Marathon. Several years before, never feeling that she fit the mold of a champion runner, she had quit running altogether. Now she's proving that it's possible to break all the rules — and wind up winning anyway.

Send your thoughts on this week's selections, suggestions and favorite reads from the section to me at Op-reads@nytimes.com. If you're enjoying this email, please send it to friends. They can sign up here.