When Humans Think…And Machines Do Too

When Humans Think…And Machines Do Too

What does it mean to think?

It’s a deceptively simple question.

We use the word think every day. We think through difficult decisions. We think about the future. Some of us think before we speak. The meaning seems obvious until someone asks us to define it.

For more than two thousand years, philosophers have wrestled with this question because they understood that our answer shapes how we understand knowledge, reason, judgment, consciousness, and ultimately ourselves. Today, that ancient question has acquired new urgency with the introduction of artificial intelligence.

Generative AI systems increasingly participate in activities we have long associated with thinking. They reason through problems, draft arguments, summarize information, generate ideas, and assist with decisions once reserved for human judgment. Whether these systems are truly thinking or merely producing convincing simulations is almost beside the point. Their existence forces us to revisit a question we may have assumed was already settled.

Or perhaps it never was.

What Do We Mean by “Thinking”?

One reason the debate feels so difficult is that there has never been a single, universally accepted definition of thinking.

For Plato, thinking was a personal experience and a journey of the soul toward the most “splendid” and “blessed” level of existence (Plato).

Aristotle approached the question differently, emphasizing that thinking relies upon imagination, perception, memory, and other corporeal activities (Gendlin).

Centuries later, Descartes grounded certainty itself in the act of thinking, arguing that while he could imagine having no body or senses, he could not separate himself from his own thinking. “I think therefore I am” (Descartes).

Kant shifted the conversation toward the human mind as the active architect of experience, proposing that the world we know is constructed by our own internal mental rules rather than being a simple reflection of an independent reality (Gardner).

Arendt distinguished among thinking, knowing, and judging, suggesting that each serves a different purpose in how we interact with the world.

These philosophers were not answering the same question.

They were asking different questions about the same phenomenon.

Psychology approaches thinking from another direction.

Rather than asking what thinking should be, psychologists have attempted to understand how thinking actually operates.

Piaget argued that children are active builders of their own knowledge who develop thinking skills through a biological process of reaching mental balance. He believed thinking moves through fixed stages, from simple physical actions to abstract logic (Müller et al, 2023). 

Lev Vygotsky emphasized that thinking is a social process, and that higher-level thinking begins between people through shared activities and language (Vygotsky, 2012).

Bruner suggested that the human mind is an “instrument for producing worlds” and that we have two main ways of thinking: one based on logic and science, and another based on stories and “narrative” (Shaffer, 1988).

Kahneman distinguished between System 1, which is the fast, automatic, and “intuitive” part of our mind that handles simple language, and System 2, which is the slow, effortful, and “logical” part of our mind that we must turn on to perform complex tasks or check for errors (Kahneman, 2011).

Cognitive science complicates the picture even further.

Simon proposed that human beings, as thinkers and actors, are actually quite simple, and that the complicated behavior we see is mostly a reflection of the complex environments we are trying to adapt to (Simon, 1996).

Hutchins challenged the assumption that thinking happens only inside individual minds by arguing that intelligence is a team effort that includes the people we work with, the tools we use, and the environment we inhabit (Hutchins,1995).

Norman suggested that technology changes cognition by changing the nature of the tasks we perform, proposing technology acts as a “cognitive artifact” or a teammate that extends our minds, allowing us to offload difficult mental work (Norman, 1993).

Clark and Chalmers extended this idea, proposing that the mind does not stop at the boundaries of skin and skull, but literally extends into the tools and environment we use (Clark & Chalmers, 1998).

None of these perspectives completely agree.

Yet together they reveal that thinking is far richer, and far more contested, than our everyday language suggests.

AI Challenges Our Answers And Our Assumptions.

Most conversations about AI begin with a familiar question:

Can AI think?

But that question quietly assumes something we may not actually know.

It assumes we already understand what thinking is.

Do we?

If thinking is reasoning…

AI appears to reason.

If thinking is language…

AI appears to use language.

If thinking is planning…

AI plans.

If thinking is solving problems…

AI solves problems.

If thinking is judgment…

The answer becomes less obvious.

If thinking requires consciousness…

The debate changes again.

If thinking requires intentionality…

The conversation shifts once more.

Perhaps AI isn’t exposing the limits of machines.

Perhaps it is exposing the limits of our own definitions.

Thinking Has Never Happened in Isolation

Another assumption deserves examination.

We often imagine thinking as something that happens entirely inside an individual’s mind.

Yet several traditions challenge this picture.

Vygotsky reminds us that thought develops through social interaction (Vygotsky, 2012).

Hutchins argues that cognition is distributed across people, tools, and environments (Hutchins, 1995).

Clark suggests that tools can become genuine extensions of cognition (Clark, 2008).

Long before AI, humans were already thinking with notebooks, maps, calculators, libraries, computers, and one another.

Artificial intelligence does not introduce the idea of distributed thinking.

It radically expands it.

The question is no longer whether humans think with tools.

We always have.

The question is whether AI represents another tool or something fundamentally different.

Is Thinking the Same as Producing Thought?

Writing, coding, explaining concepts, proposing solutions, generating arguments, are all normally associated with thinking.

But AI can do all of that.

So are the products of thinking the same thing as thinking itself?

Consider two possibilities.

One possibility is that thinking is best understood through its observable outputs. If so, highly capable AI systems may deserve to be regarded as thinkers.

Another possibility is that thinking involves something beyond observable behavior, like consciousness, intentionality, understanding, moral judgment, or lived experience.

If that is true, AI may be performing activities that resemble thinking without participating in thinking itself.

The answer depends less on AI than on what we ultimately decide thinking is.

Why This Question Matters

At first glance, this may sound like an abstract philosophical debate.

It isn’t.

Our answer will shape how we educate children.

How we define expertise.

Evaluate professional judgment.

Design workplaces.

Cultivate creativity.

Assign responsibility.

And even understand ourselves.

Questions that once belonged primarily to philosophy now have practical consequences across nearly every domain of society.

Where the Inquiry Begins

Does AI think? I don’t have the answer to that.

I’m not convinced we’ve developed a definition of thinking precise enough to answer the question with confidence.

What I do know is this.

Artificial intelligence has transformed one of philosophy’s oldest questions into one of society’s most urgent ones.

Before we ask whether machines think, or whether AI will replace human cognition, we should begin with a simpler, yet much harder, question.

What does it mean to think?

Bibliography

Arendt, H. (1978). The life of the mind (M. McCarthy, Ed.). Harcourt, Inc.

Clark, A. (2008). Supersizing the mind: Embodiment, action, and cognitive extension. Oxford University Press.

Clark, A., & Chalmers, D. J. (1998). The extended mind. Analysis, 58(1), 7–19.

Descartes, R. (2008). Meditations on first philosophy: With selections from the objections and replies (M. Moriarty, Trans.). Oxford University Press. (Original work published 1641).

Gardner, S. (1999). Kant and the Critique of Pure Reason. Routledge.

Hutchins, E. (1995). Cognition in the wild. MIT Press.

Kahneman, D. (2011). Thinking, fast and slow (1st ed.). Farrar, Straus and Giroux.

Müller, U., Ten Eycke, K., & Baker, L. (n.d.). (2023). Piaget’s theory of intelligence. In Handbook of intelligence: Evolutionary theory, historical perspective, and current concepts. (Presented by Bowen Xu).

Norman, D. A. (1993). Things that make us smart: Defending human attributes in the age of the machine. Addison-Wesley.

Norman, D. (2013). The design of everyday things (Revised and expanded ed.). Basic Books.

Plato. (1987). The Republic (D. Lee, Trans.; 2nd ed. rev.). Penguin Books. (Original work published ca. 375 B.C.E.).

Simon, H. A. (1996). The sciences of the artificial (3rd ed.). MIT Press.

Shaffer, T. L. (1988). Actual minds, possible worlds [Review of the book Actual minds, possible worlds, by J. Bruner]. The American Journal of Jurisprudence, 33, 241–250.

Vygotsky, L. S. (2012). Thought and language (Rev. & expanded ed.; E. Hanfmann, G. Vakar, & A. Kozulin, Eds. & Trans.). MIT Press. (Original work published 1934).