AI: Can machines truly think?

Artificial intelligence is all the rage right now, promising to change many aspects of life. Its rise is akin to Pandora’s box, and we can only guess what this technology has in store.

For some, it will be the greatest thing since sliced bread will release humanity from many more mundane aspects of work. For others, the prospect of AI proliferation is seen as akin to self-suicide for the species.

But should we be worried? Could it ever really replace humans? Can it, or could it ever think like us? Let’s take a look. Before we get into the weeds, it might be worthwhile to explain what thinking is.

What is thinking?

You do it every second of every day, but what exactly is “ thinking ?” Sadly, we don’t know for certain, but it is a kind of “you’ll know it when you see it” affair.

Fittingly enough, many of our greatest minds have spent hours thinking about what thinking is. From Plato to Aristotle to more modern philosophers, what is going on in our heads has fascinated and perplexed us for millennia.

While many theories have been proposed, we have discovered that it is a cognitive process (obviously) that can occur without sensory stimulation. Depending on your definition of “thinking,” it could be something unique to humans or something shared by other animals (albeit to a much lower degree than human beings).

Whatever the case, AI is being designed to attempt to mimic our remarkable capacity for thought, so we’ll stick with how humans think. Thinking tends to involve some form of problem-solving, such as judging, reasoning, concept formulation, and deliberation.

Others may also include memory or imagination, but this is not universally accepted as these processes can be argued to require sensory input like sight, sound, etc. Other schools of thought, like functionalism and computationalism, posit that mental states are defined by function rather than composition.

In this sense, our brains may function similarly to squishy soft computers, which could mean that AI could, one day, actually think in a true sense. Many other theories would require a book’s length to cover, but we think you get the idea.

How does AI work?

AI systems like large language models (LLMs) and art generators function by processing vast amounts of pre-existing data to simulate human-like responses or creative outputs. LLMs, in particular, are trained on diverse text data and learn language, grammar, and context patterns.

When given a prompt, they generate coherent text by predicting the most likely word sequence based on their absorbed data. However, they do not “think” or understand the meaning (as it were); their responses are statistical outputs rather than original ideas.

AI art generators, like those used in image creation models, operate in pretty much the same way. Put simply, they can analyze millions of visual inputs (such as photos, paintings, and illustrations) and learn to recognize patterns in style, form, and color.

As with LLMs, when prompted (noticing a pattern yourself?), they generate new images by blending these patterns. By doing this, they effectively remix the elements they’ve been trained on.

In fairness, this is not dissimilar to apprentice artists or even seasoned artists who take inspiration from other works to make their own. However, true artists (especially masters) find their style, continue to hone their trade, and can, in some cases, create some of the most beautiful pieces of art the world has ever seen.

Both AI types excel at mimicry and simulation but cannot experiment, understand, or create with genuine intent. They imitate creativity, drawing from human-made inputs without the subjective experience that drives true human intelligence and invention.

They also require human input to get the ball rolling. Without human input, the AI will sit there and not do anything. AI tools are collaborative, with humans giving the initial seed of an idea and AI doing the rest.

And that really could be the Achilles Heel for any wannabe bone fide AI.

Imitation is the sincerest form of flattery

The key point is whether there is something inherently unique to living creatures (humans or otherwise) and how their brains process data compared to a computer—at least for now.

Some experts believe that AI may achieve “ artificial general intelligence ” (AGI), enabling it to perform any intellectual task that a human can. However, others argue that true consciousness and understanding might be beyond AI’s reach because it lacks a subjective point of view.

In this sense, it will only ever be very convincing, but ultimately just a sophisticated, high-fidelity illusion of intelligence—a complex human-input-in-computer-generated-output machine, if you will.

This is because any “creations” produced by AI are essentially pattern-based outputs generated from human-made inputs or “inspirations” (aka large data sets). Although they may imitate creativity, it can be argued that AI will always be limited by the data they have been trained on and cannot make imaginative leaps like humans.

True creativity in humans often stems from emotional or existential triggers, such as the fear of death or the desire for a legacy. In this sense, AI will always be a highly advanced tool for simulation, but true experimentation—where outcomes are unknown, and stakes are real—remains a uniquely human trait.

Adapt or die

The main reason is AI’s current and presumed inability to conceive of subjectivity. In its truest form, subjectivity is defined as a quality of being based (or influenced) by our feelings, tastes, experiences, and opinions.

In humans, this is the product of millions of years of evolution and arguably an adaptation for survival on this planet. Whatever it is, it has gifted humans with the ability to play and experiment to find novel ways to survive and flourish.

However, even with superior processing power, AI lacks the innate sense of urgency or existential reflection humans develop through lived experience. Thinking, in other words, might require mortality, or at least the very real possibility of dying if your decision is wrong.

Adapt or die, if you will.

Since machines or computer programs can never truly “die,” they will always lack this vital component of thought and will never really be able to “think” as we understand it. If true, it will never be able to create, only copy or remix the creations of humans.

Human-AI unification on the cards

That doesn’t mean it won’t remain a powerful tool to augment true intelligence. It is highly unlikely that AI will now go the way of the dinosaurs.

If this is AI’s destiny, as others have suggested ( like Elon Musk ), the future may see some unification between man and AI to get the best of both worlds. We are already seeing the early stages of AI being integrated into existing products like search engines or computer operating systems.

In the future, we will likely only see more integration into other aspects of our lives, such as our cars or home appliances, and it is not inconceivable for ways to be found to speed up human-to-AI interaction. Elon Musk’s Neuralink is one example.

And that is your lot for today.

Humans may be in trouble if AI can make the quantum leap from a highly advanced task-completing tool to a self-prompting and initiating tool. However, as we’ve argued, this may not be possible given the nature of machines and AI.

But it is important to note that we are still in the early stages of this technology. Who knows what is around the corner?

But what do you think?