Trump Orders Agencies to Drop ‘AI’ for ‘Super Intelligence’

It was William Shakespeare who penned that famous locution: “That which we call a rose / By any other name would smell as sweet.” One can forgive the playwright — after all, he placed the words in the mouth of Juliet, who was desperate to prove that the name of her beloved had nothing to do with his nature. Charity suggests that we assume the point was literary rather than a statement of belief.

Despite Juliet’s assurances, names communicate something about reality. It was Shakespeare’s countryman G. K. Chesterton who observed: “When somebody wishes to wage a social war against what all normal people have regarded as a social decency, the very first thing he does is to find some artificial term that shall sound relatively decent.” Later, George Orwell pointed out in his essay “Politics and the English Language” that words — especially idioms and euphemisms — are so powerful that they shape the way we think.

Such terms might include “euthanasia,” which means “good death” and refers to an end that is anything but good, or the phrase “a woman’s right to choose,” which we apply to women who choose the cubicle but not to those who prefer a kitchen and the blessings of many children. Then, there’s the term “artificial intelligence,” which describes something that is certainly artificial but not at all intelligent.

Or, at least, we called it artificial intelligence until Tuesday, when Donald Trump signed an executive order directing the executive branch to call it “super intelligence,” a euphemism even worse than the one it replaces.

Trump’s decision to rename AI — at least as far as the executive branch is concerned — is not necessarily a bad move, regardless of how dishonest it is. In recent months, the technology has received a bad rap, and so protesters now gather outside tech conferences and carry signs reading things like “AI rots your brain” and “eat the rich, save the planet.”

The concerns are numerous, vary in their merit, and are mostly unrelated. Some people are worried that the infrastructure necessary to sustain the AI revolution will destroy the planet. Others find AI slop disconcerting and worry about its impacts on art, writing, and culture. Still others are concerned that their jobs will be eliminated by hordes of robotic agents that don’t require a salary. Following the assertion of an ex-Anthropic employee that “the people building AI earnestly believe that it could kill us all by the end of the decade,” many people believe the apocalypse is near.

Given that the United States is in a technology race with China, none of these opinions — regardless of their validity — are helpful to Trump, who is trying to win that race. So he decided to host an “AI Day” at the White House on Tuesday with the goal of alleviating the public’s concerns. Tech oligarchs signed a voluntary accord to establish additional guardrails on the technology. The White House announced the launch of America.gov, a website with a built-in chatbot whose job is to connect users with the right forms to accomplish tasks like applying for passports, getting a replacement Social Security card, or changing their mailing address with the USPS. And, of course, Trump signed his executive order renaming artificial intelligence.

As far as the president was concerned, it was about time for a rebranding. “The capabilities of today’s frontier systems do much more than imitate or automate discrete aspects of human intelligence,” the executive order reads. “As these capabilities continue to improve, they increasingly represent not merely artificial intelligence, but a new era of Super Intelligence.”

There are two problems with the term “super intelligence.” The first is that it already refers to something in the tech field: namely, machines that are capable of outperforming humans cognitively. As Politico pointed out, we haven’t technically made a superintelligent machine yet.

The second problem is that large language models (LLMs) aren’t intelligent. Chatbots work by predicting the most likely next word in a sentence based on neural networks trained on vast quantities of text. Intelligence is the ability to understand and learn new things. While it’s true that Claude and ChatGPT are good at constructing sentences and paragraphs (arguably better than some humans), their ability to independently comprehend things, come up with new ideas, or link concepts in novel ways has not been demonstrated. And yet, even your average seven-year-old is capable of those things.

In short, LLMs merely have the appearance of intelligence.

One could forgive the term when “artificial” qualified it. It’s harder to do so when the euphemism is so totally obscuring that it ceases to describe reality at all. To be “super intelligent,” machines would have to demonstrate that they can know and understand things better than the human mind does — a fantasy of the transhumanists.

Transhumanists, in the words of theologian Paul Scherz, “imagine that AI will actually surpass human intelligence and that our destiny is to merge with these tools, to become our machines, to bow to that which is superior because more efficient and powerful in their eyes.”

Whether the White House intends to play to these fantasies or not, choosing to employ the term carries certain implications. At the very least, it opens the door to acclimating the public to the idea that perhaps the marriage of machine and man isn’t all bad.

This isn’t to argue that we should regulate the tech industry. It also isn’t to insist that the advent of chatbots heralds the apocalypse — after all, the Trump administration is right to observe that these technologies can be incredibly helpful in our day-to-day activities. Opposing them is about as futile as opposing the steam engine in the age of the transcontinental railroad.

It’s simply to observe that, in naming our new creations, we should at least be honest with ourselves. We are, after all, merely human, and our machines are merely machines.