Here's More Evidence That Trump Is an Algorithm


Donald Trump has been ridiculed for responding to the bad news on Covid-19 by suggesting that the U.S. is simply conducting too many tests. “Without testing, or weak testing, we would be showing almost no cases,” he complained. Or as he put it at a recent news conference: “When you test, you create cases.”

Actually, such statements offer a valuable insight into the way the U.S. president’s mind works.

Of course, testing doesn’t create sick people, it merely discovers them. If there were no sick people to discover, testing would not create bad news. So it’s logical to dismiss Trump’s absurd reasoning out of hand. I did that for weeks. But then I started thinking, why not take this seriously?

He has repeated the point enough to suggest that it deeply resonates with him, which things rarely do. He’s usually capable of changing his mind quickly, especially when roundly ridiculed. But for some reason, this absurdity has stuck.

Moreover, he’s kind of right in a shallow way. If we didn’t count coronavirus cases or deaths, then we’d have less specific bad news to talk about. It’s childlike logic, but it’s consistent. Don’t collect bad information about me or the state of the world, and everything will be better and people will stop picking on me.

It’s akin to Trump’s obsession with TV ratings, with no regard for the substance of what generates them. In the midst of a worldwide pandemic, economic depression and hugely consequential presidential campaign, Trump’s volley is to suggest that his opponent, Joe Biden, can’t garner enough viewership to keep America great.

All this indicates that Trump has a very simplistic model of the world, focused on data -- ratings, case counts, polls –- but utterly independent of reality, duty, guilt or conscience. If the numbers look good, then things are good. If the numbers look bad, then change the numbers or use different numbers.

The narrowness of Trump’s focus supports a favorite theory of mine: that he can best be understood as a form of artificial intelligence, a machine-learning algorithm. In the game of checkers, for example, such an algorithm might assign each configuration a value and make moves to maximize that value. The target metrics must be kept simple, typically too simple to capture the complexity of the real universe. This works fine for checkers.

Problem is, Trump isn’t playing checkers. He’s the leader of what remains the world’s most powerful nation, playing a game with global consequences, and his preferred measure of success doesn’t capture even a sliver of what’s really going on. If only he could be reprogrammed to understand that.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

Cathy O’Neil is a Bloomberg Opinion columnist. She is a mathematician who has worked as a professor, hedge-fund analyst and data scientist. She founded ORCAA, an algorithmic auditing company, and is the author of “Weapons of Math Destruction.”

©2020 Bloomberg L.P.

BQ Install

Bloomberg Quint

Add BloombergQuint App to Home screen.