This “logic” is naturally extended to a claim that computing is just like thinking. Nothing could be further from the truth and in fact, everybody intuitively understands that.

Let’s take the title of the article to see how this deception starts. “Thinking computer” seems like a reasonable claim if the mind is like a computer as the popular culture keeps telling us. However, as soon as I replace the “computer” with “machine” the claim that “machine thinks or cares” becomes obviously ridiculous. By definition machines do mechanical work, they don’t think, show emotions or “act” in any way similar to human beings. They don’t act at all unless turned on with a switch. So a “thinking computer” or “brain is like a computer” are just metaphors that are useful in some circumstances but cannot be taken literally.

What does a computer do? There is a very simple definition for it. It’s a process where a computer:

1. gets input from the external world
2. applies a series of algorithms on it
3. produces an output

But thinking is not like computing. For a person to think means to “contemplate” a situation. Thinking in itself doesn’t require input, an algorithm or an output. In its basic form, it can be summed up in 3 Latin words according to Descartes:

Cogito, ergo sum = I think, therefore I am

In short, thinking is nothing like a computer’s calculations.

AI shows signs of intelligence, but it’s not intelligent. It’s like a super-speedy global librarian who knows where to find everything ever recorded and how to use it. But it doesn’t know how to invent anything worthwhile.

I will cite only one source for this but it should be authoritative enough to prove my point, and it’s a recent, extensive Apple study by its top researchers to find out how much “thinking” goes on in AI.

https://www.linkedin.com/pulse/illusion-thinking-what-apples-ai-research-reveals-humanai-grobben-jjgzc/.

The Apple study shows that AI doesn’t “think” independently for itself; it can only rehash the existing stuff that has been published, whether the stuff is true or false. The so-called “reasoning” versions do better, but not good enough.

AI cannot even rehash in its own way existing sentences unless these have already been rehashed by others and published. For a little-known subject, even reformatting a paragraph is an impossible task for AI. I know this from personal experience editing my son’s research work.

I had no problem reformatting/rephrasing my son’s medical and biological research papers, even though I didn’t fully understand them due to my limited understanding of the subject but I understood them enough to be able to edit them.

So what is a human being if not a machine? The key concept to understanding “human intelligence” is actually “understanding,” as I mentioned in my previous blog. The philosophers have known this for decades; the techies still don’t get it; to manipulate symbols such as human language, speech, and writing is one thing, but to understand these symbols is something else.

Let’s end on a positive note by defining what is the essense of a person so that we can show that this is not simply a denial of AI abilities but a reasoned adjudication of the AI shortcomings compared to human capabilities.

It’s actually quite simple. Since only people are generally “understood” to understand, it’s safe to say that the real determinant of understanding is consciousness, as only people are conscious. Well, let’s qualify that, linguistic consciousness, as I do believe that animals and even plants (they can be put to sleep by anesthetics) have consciousness, and if you have a cat, a dog, or a rabbit, you see it every day. Another aspect that accompanies consciousness is free will, or the will. To be free, one must understand; without understanding, there can be no freedom of action. And to understand, one must be conscious; without consciousness, there is no understanding.