# The Artifact of Thought > Source: > Published: 2026-08-17 17:49:16+00:00 ###### [Artificial Intelligence](/us/basics/artificial-intelligence) # The Artifact of Thought ## Is AI driving a disconnect between the poem and the poet? Posted August 17, 2026 [ Reviewed by Monica Vilhauer Ph.D. ](/us/docs/editorial-process) ### Key points - Writing was long treated as an artifact proving a mind behind it. - AI now creates the same artifacts from a different architecture. - This forces us to question how we recognize thought itself. It's really rather simple. A poem once seemed to imply a poet. A mathematical proof implied a mathematician. An elegant argument suggested someone doing the arguing. But now, it's getting complicated by [artificial intelligence](https://www.psychologytoday.com/us/basics/artificial-intelligence). We can now encounter the artifact of thought without knowing what kind of cognitive architecture (if any) produced what we recognize as thinking. I want to focus on the word artifact. Recently, I used it in a casual conversation about [education](https://www.psychologytoday.com/us/basics/education). We were talking about writing and learning, and I described a student’s essay as an artifact or proxy of learning. That seemed straightforward and carried the stamp of logic. We can’t actually see learning, so we look at what the student produces and make some reasonable judgment about what happened in that student’s mind. And in that context, we craft that into a single letter grade. The conversation stopped, but my thinking continued. If an essay can be an artifact of learning, perhaps writing itself can be understood as an artifact of thought. And that creates a much more interesting problem. ## Thoughts and the Thinker Thought is surprisingly difficult to get our hands around. I can tell you what I’m thinking and even work through a problem in front of you with a good old fashioned pen and paper. But the thought, yours and mind, remains curiously inaccessible. What you encounter is its expression—a derivative of sorts. And from that expression you reason backward toward the [cognition](https://www.psychologytoday.com/us/basics/cognition) that produced it. Writing makes this particularly clear. When I read a beautiful paragraph, I almost automatically imagine a person or even a mind behind it. Someone selected the words, seeks the rhythm and probably shifts direction a few times. Maybe the writer struggled with a sentence or discovered halfway through that the original idea wasn’t quite right. I don’t consciously reconstruct any of this. I simply assume the words carry some trace of the thinking that produced them. The same thing happens all the time, from a mathematical proof to a clever joke. We embrace the result and form some impression of the [intelligence](https://www.psychologytoday.com/us/basics/intelligence) behind it. There’s a lot packed into that little cognitive leap, and until recently there wasn’t much reason to examine it and we just left it alone. But today, AI gives us a reason. ## Recognizing Another Mind Now, I can encounter an extraordinary paragraph without knowing if someone struggled with it for days or if it just appeared on a screen with the mechanical click of a button. Its quality and meaning can still be there but it becomes harder to reconstruct the cognitive story that preceded the paragraph. So, here's the question, and one that might be essential for our times. How do we know when someone else is thinking? When another person speaks or writes, I can recognize something that seems connected to my own interior experience. Someone says something insightful and I make an inference about the person who said it. She’s smart. He understands the problem. That’s a great idea. Language becomes evidence of another mind because we assume some relationship between the expression and the cognitive process that lives behind it. AI introduces another source of language into that equation, and its architecture is [profoundly different](https://www.psychologytoday.com/us/blog/the-digital-self/202504/how-humans-think-and-ai-generates) from ours. It can produce an observation that surprises us or craft an argument that changes the direction of a conversation. Whatever we ultimately decide about whether AI thinks or understands, something consequential has already happened and it demands our [attention](https://www.psychologytoday.com/us/basics/attention). Artifacts we associate with cognition can now emerge from something very different from a human brain. ## AI as a Cognitive Contrast Agent Maybe AI is a sort of cognitive contrast agent. In medicine, a contrast agent helps reveal structures that were already there but difficult to see such as blood vessels. AI may be doing something similar to our understanding of thought. By giving us another architecture capable of producing artifacts that overlap with our own, it begins to expose assumptions that were difficult to notice when human cognition was essentially our only reference point. We see intelligent behavior and infer intelligence. We encounter a meaningful sentence and infer understanding. We see something creative and imagine [creativity](https://www.psychologytoday.com/us/basics/creativity) somewhere behind it. These inferences feel natural because they were built around the human example. For a long time, there was little reason to separate the artifact from the presumed cognitive process behind it. AI is changing this as it puts pressure on that relationship because the same outward signal can now have a very different history. This is also why the argument over whether AI “really thinks” may be too narrow. We don’t need to settle that question to recognize what has already changed. AI has complicated the relationship between intelligent expression and the assumptions we make about its source. And perhaps this is the part I find most interesting. A poem can still move me. A proof can still strike me as brilliant. An argument can still change the way I see something. The artifact hasn’t necessarily lost its power. It's become less revealing about the process that brought it into existence. ## What Was the Artifact Telling Us? We have never had direct access to another person’s thought. We've had all sort of "stuff" that lives at a distance. Words, gestures, ideas, creations are the traces that minds leave in the world. From those traces, we constructed our understanding of other minds. And I think we became so accustomed to making that inference that we rarely noticed we were making it at all. AI gives us a new kind of trace with a very different origin. And suddenly something that once seemed fairly obvious becomes less so. A beautiful sentence remains a beautiful sentence. But what exactly does its existence allow us to conclude about what happened before it appeared? Maybe AI will eventually tell us something profound about artificial intelligence. I’m beginning to suspect that one of its more interesting contributions may be what it forces us to reconsider about our human cognition. The artifact is still right there in front of us. What lies behind it has become much harder to locate, and I’m not sure we (yet) understand what that does to the way we recognize thought when we see it.