Why are we asking "How can we simulate the brain?"

Question

As I've thought about AI, and what I understand of the problems that we face in the creation of it, I've noticed a recurring pattern: we always seem to be asking ourselves, "how can we better simulate the brain?"

Why are we so fascinated with simulating it? Isn't our goal to create intelligence, not create intelligence in a specific medium? Isn't growing and sustaining living brains more in line with our goals, albeit a bit of an ethical controversy?

Why is this exchange's description: "For people interested in conceptual questions about life and challenges in a world where 'cognitive' functions can be mimicked in a purely digital environment?"

To condense these feelings in a more concise question: Why are we trying to create AI in a computer?

Answer 1

Human intelligence is very general / broad in its scope. This is self-evident, and whatever AI ends up to be, we'd like it to be a general problem solver as well (cf. Simon and Newell). Taking liberal interpretations of your question...

Why AI in a computer?

Computers, to the extent that we can frame problems in general as a solvable computational problem, are also general problem solvers. Wether this is actually the case (can you compute meaning or feels?) is up for debate (cf. computational functionalism, hyper-computation), but it is part of the artificial intelligence project to make a claim on this statement.

Why do we think a computational framework brings us any closer to an understanding of cognition / consciousness?

Good question, and frankly there is no good answer to that asides from "its the best thing we got".

TL;DR "computational functionalism", a lot of the literature in psychology and philosophy seems to converge towards an understanding of cognition as "computational" (as in information processing: the V1 stream in the brain processes "early visual information") and functional (goal directed grounded on "meaning", ex: "i scratch itch because itchy", as opposed to "i am moving atoms").

However the two theories don't mesh together well (cf. Chinese Room Argument, and the many other arguments in a similar flavour) despite their independent successes in the theory of mind. Why this is the case nobody quite knows...

Why not AI in something that isnt a computer?

I don't know, but to the extent that our understanding of the world is grounded in math, then it being in a computer is sufficient anyways.

Maybe there are other paradigms of understanding the world though. Fingers crossed 🙏

Why are we asking, “How can we simulate the brain?”

Because its the best tentative understanding we have of an "intelligent faculty", though it should be noted that various methods in machine learning don't seem to be directly inspired by biological implementation (kNN, statistical methods, as opposed to neural nets)

Further reading: http://www.scaruffi.com/nature/mach01.html

Answer 2

I don’t think AI is simulating the brain functions and not even close. Do you know how the nervous system work? How the neutrons transmit signals with action potential? Pathway analysis? Splicing junctions?

AI is not about simulating the brain at all. We don’t simulate the biology pathway, we don’t simulate alternative splicing, we don’t have proteins in our models.

Instead, AI is a field with tons of mathematics. You give some data and try to extract complicated non linear pattern.

Answer 3

There are a number of reasons why a simulated brain might be better than creating a real brain. One reason is computers can live indefinitely (kind of). Brains may not be able to live forever and there might not be a way to transfer information from one brain to another. One of the principle advantages of a computer then is that it could have more experience than any brain could have in its lifetime. Another reason is that there are a lot of things we don't know about the brain. Even if we were able to replicate the brain we would have a hard time using it in the way that we want until we fully understand it. The simulated brain doesn't have this problem. We know exactly how artificial neural networks develop, and thus there is not as much that we don't understand.

Those answers tell you why we might want a digital brain, but your question seems to also ask why study the digital brain over a biological brain? This seems to imply that we can't do both, but in fact there are many research groups doing work in areas that contribute to growing living brains (Max Planck Institute of Molecular Cell Biology and Genetic (MPI-CBG), the Medical Research Center in the UK, etc.).

Answer 4

I think a worthwhile extension of this line of thought is "why not both?"

I do not believe there is anything preventing approaching the problem from both sides at once. There is a great deal of research on both sides (biological research and computational research), but considerably less on the integration of the two (although there certainly is some, such as in the development of modern prosthetics that allow some degree of control).

Given the adaptability of the human brain in terms of adjusting its own structure, the most expedient approach may be to consider what it would take to create a non-biological medium that biological neurons could interface with sufficiently to essentially "program" them in the same manner it does when repairing itself with biological neurons. Leave the hard work to the thing that already has the blueprint. Or in other words, the Ship of Theseus but with brain cells.

Not that such a task would be anything close to approaching simple or easy, given our still lacking knowledge of neurological structures and the difficulties in getting a non-biological interface that is capable of the required sort of communications and adjustments that biological neurons can have performed and on a size scale that would be practical.

I wish I could point to some research related to this, but I don't know about any specific research papers, although I know it's not a completely untouched upon subject.

Answer 5

For what its worth (and having done a bit of study on this and being really interested in the topic): the answer seems to go back to the beginnings of AI and even earlier (Turing's 1936 paper in which he introduces what's now called the Turing machine).

John McCarthy's filer for the 1956 Dartmouth College summer workshop on "Artificial Intelligence" (which name introduced the term "Artificial Intelligence") in part says:

"The study [workshop] is to proceed on the basis of the conjecture that every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it."

This references Turing's 1936 paper where a machine or natural system is described, and the description is run in a computer. To simulate is to quite precisely describe a system then run the description (transformed a bit – but the result is still a description) in a computer. The description is the program. The description needs to be precise, as indicated in the Church-Turing thesis.

So the idea of simulation is core to the computational theory of what a digital computer can or might do. So it's also core to the computational theory of mind (the organic brain being a natural system), and hence to AI.

That said, it's obviously a crazy idea to try to quite precisely describe the organic machine that is a human brain. I mean how many neurons? 100 billion. Quite precisely describe each and every single one of these, and each and every of the up to 10,000 connections that connect to each and every single neuron. Crazy with a capital C. And to suppose there are degrees of simulation of the brain, or that the mind is somehow a simplification of the brain, or that the description can be in higher level concepts, not neurological ones, is just to admit that the description is not quite precise. An adequate simulation of a brain would be terribly detailed.

So why do we hear so much about AI trying to simulate the brain? Answer: AI has no other word to express what it does.

In my view, AI ought to be trying to work out the data-processing principles of the organic brain, not trying to describe the causation of the brain. AI doesn't know the principles of perception or the principles of general knowledge. It's incredible to say this – seeing as both are so absolutely fundamental to human intelligence. But AI doesn't know the principles. It ought to be trying to work them out. Then – once discovered – to work out how these principles could be realised in a computer.

You suggest that there's a binary choice between AI trying to get a computer to simulate the organic brain, and trying to grow organic brains in a dish. But there's actually a third option. Computer can do things other than simulate (i.e., other than compute). Maybe these other things might include embodying the principles of organic brains.

There are two really big areas here: (1) what are the principles of intelligence? (2) what are the non-computational things computers can do?

You ask why AI is concerned with the digital environment rather than, say, growing organic brains in a vat. But AI is basically an engineering project (building something with a designed causality) and even though AI knows only a little about the causality of what it's trying to build, the digital computer seems to be the only viable platform, at present, with enough individually addressable memory locations and processor speed to cope with semantic structures that would result from an adequate sensory interaction with the environment.

Answer 6

Being the OP, I have already put some thought into this question.

I think that computers are an attractive medium for simple AI because they are easily available and researchers are already familiar with them. In addition, science fiction writers of the last century were hopeful of the capabilities of computers and placed in our culture a dream of computer AI.

But I also feel that perhaps other less explored fields would be better suited to the creation of strong AI. In particular, thinking about the nature of biology excites me. But as I understand it, we still know so little about how it biology works, let alone how to control it. But I feel this is where we should be focusing.

Researchers know that current computing hardware has limitations. GPUs are better suited than CPUs. Some CPUs have new hardware designed for AI computations. I suspect that this realization of the inadequacy of conventional hardware will continue until our hardware is nearly identical to the biology we are trying to simulate. After all, what simulation could ever be better than what it is trying to simulate?