1
$\begingroup$

There is a Wikipedia page on brain simulation, but:

Question: Could there be (experimental) evidence that the (human) brain cannot be simulated by a computer?

Note that the Wikipedia page Mind uploading has a section called Practical issues, but Wikipedia says that this section needs expansion. It is followed by sections on philosophical, ethical, legal or political issues.

My purpose is not philosophical here. Roughly speaking, I just want to know whether there are research works (published in serious peer-review journals) showing that the human brain cannot be simulated by a classical computer.


Previous version of the post

Previous title: Can the human brain be reduced to a biological AI?

The question is in the title. Let me clarify what I mean by « biological AI » : it looks like an oxymoron if you understand biological as natural and artificial as non-natural, but let understand biological as organic and AI as what it actually does (and not as what it is made of). On the other hand, I am (obviously) not asking whether the human brain was designed...
AI is here understood in its current classical sense, and not in what it could be in the future with (for example) quantum computers, so that if one can shows (for example) that the (humain) brain involves some quantum processes which cannot be simulated classically, then the answer should be no. Note that the quantum processes were just used here as an example, the question does not reduce to the existence of such processes.

$\endgroup$
3
  • 3
    $\begingroup$ I really had hard time reading your question, Simple processes are often challenging to approximate using AI, For example radio-active decay cant be approximated usng AI and they are sporadic processes, other examples are thermal fluctuations at micro level. AI is capabile of learning those processes that show some trend like you can see facial recognition algoritms and even facial reconstruction algos. I doubt the comlexity of human brain can be approximated using AI (although simple tasks have been approximated like word prediction etc) $\endgroup$
    – gfdsal
    Jan 24 at 21:05
  • 1
    $\begingroup$ Possibly related: psychology.stackexchange.com/questions/18591/… $\endgroup$
    – Fizz
    Jan 26 at 13:31
  • 1
    $\begingroup$ Possible duplicate of Can humans be reduced to a function? Also related: When can we say that machines are conscious? and Neural Microfilaments for Computation? $\endgroup$
    – Arnon Weinberg
    Jan 27 at 2:13
1
$\begingroup$

To my knowledge (I hold an advanced degree in the field), there is no widely-accepted research that documents that the brain cannot be simulated using a classical computer.

Our current understanding of the human brain is that the primary functional mechanisms consist of neurons that transmit data via chemical and electrical connections.

That said, we have no comprehensive, finite, and accurate definition of "consciousness", and therefore we cannot prove or disprove consciousness of a computer simulation. For that matter, as much as we have tried, it is questionable if we can truly prove or disprove consciousness in human beings, although that is the topic of much debate. It largely depends on how one defines the term "consciousness".

Outside of consciousness and several other more abstract and philosophical matters (such as feelings and emotions), we know of no limits to our cognitive simulations.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.