President Donald Trump, by executive order, has changed the language used by the federal government from “Artificial Intelligence” to “Super Intelligence.” On the surface, this may seem like just a change in terminology. But I believe that the word “super” carries with it an assumption that may be more consequential than just a name change.

A few months ago, I wrote a post titled “We’re Measuring AI on the Wrong Ruler.” My argument was rather simple. When we suggest that AI is smarter than humans, we have already made an assumption that human cognition and machine intelligence can be measured on the same line. In this context, we begin to imagine intelligence itself on a single continuum, a bit like measuring it on a number line or axis.

There’s something historically familiar about the word super, and it’s a good place to start. In the early days of nuclear weapons, the hydrogen bomb was coined “the Super.” This weapon was understood to be vastly more powerful than the atomic bomb. And the term made sense because the explosive yield was measured using the same scale. More energy meant a more powerful bomb. But I’m not sure intelligence works that way.

Calling AI “superintelligent” leverages a similar word but hardly measures something as objective as kilotons of TNT.

So, this new language of superintelligence may take this assumption into new territory. “Super” establishes a hierarchy where AI is moving up along the same “line” that measures human intelligence. In this structure, it can occupy a position above us. But I’m uncertain (and uncomfortable) that this is the right way to understand what’s actually happening. Lumping artificial computation with human cognition is, to say the least, an oversimplification of both.

I’ve written about Anti-Intelligence as a term to discuss the possibility that AI represents a fundamentally different cognitive architecture that might force us to reconsider the “single line” perspective. And without getting into the complexities of quadrants and orthogonality, here’s the key point: AI can generate remarkable ideas without having the lived experience of human thought.

I believe this difference becomes more important as AI becomes more powerful. But performance and wisdom are very different things. Being very good at solving problems doesn’t automatically grant AI newfound authority. In fact, it’s essential that we recognize that computational power doesn’t confer moral authority even when AI is placed above human capacity.

Here’s where the language itself becomes interesting. If we begin describing AI as superintelligent, we may begin to change our relationship with it. If AI is positioned “above us” intellectually, deference to it might just follow along. Why struggle with a problem ourselves? Why challenge AI’s answers when it’s “linguistically validated” to be our cognitive superior? The “super” label may even transform “super” capability into a sort of intellectual status like the new Ivy League or the better Nobel Prize. And don’t forget, intellectual status often shares a border with authority (or finds its way there).

To me, the critical issue isn’t simply if “Super Intelligence” is a better name for AI. Beyond that, we need to ask what happens to our human cognition when we begin structuring our relationship with “a machine” that has advanced past us on that cognitive number line.

So, maybe we are still using the wrong ruler. And if we place AI at the top of that ruler and call it superintelligent, we need to be very careful about our assumptions on all that rests below—particularly ourselves.