Nautilus

The Fundamental Limits of Machine Learning

A few months ago, my aunt sent her colleagues an email with the subject, “Math Problem! What is the answer?” It contained a deceptively simple puzzle:

She thought her solution was obvious. Her colleagues, though, were sure their solution was correct—and the two didn’t match. Was the problem with one of their answers, or with the puzzle itself?

My aunt and her colleagues had stumbled across a fundamental problem in machine learning, the study of computers that learn. Almost all of the learning we expect our computers to do—and much of the learning we ourselves do —is about reducing information to underlying patterns, which can then be used to infer the unknown. Her puzzle was no different.

As a human, the challenge is to find any pattern at all. Of course, we have intuitions that limit our guesses. But computers have no such intuitions. From a computer

You're reading a preview, sign up to read more.

More from Nautilus

Nautilus5 min readScience
In Brain’s Electrical Ripples, Markers for Memories Appear
Reprinted with permission from Quanta Magazine’s Abstractions blog. It’s very easy to break things in biology,” said Loren Frank, a neuroscientist at the University of California, San Francisco. “It’s really hard to make them work better.” Yet agains
Nautilus17 min readTech
Who Will Design the Future?: AI will be staggeringly diverse. Its developers should be, too.
Ada Lovelace was an English mathematician who lived in the first half of the 19th century. (She was also the daughter of the poet Lord Byron, who invited Mary Shelley to his house in Geneva for a weekend of merriment and a challenge to write a ghost
Nautilus11 min readScience
A Novelist Teaches Herself Physics: To explore loss and mystery, Nell Freudenberger journeyed into the atomic world.
Helen Clapp, a professor of theoretical physics at MIT, recounted the biggest news of 21st century physics, the detection of gravitational waves by the Laser Interferometer Gravitational-Wave Observatory (LIGO), an international collaboration of scie