> The analogy that brains are like machines is harmful because it makes it sound like there is some operating manual and simple cause effect relationships that are known.
Why? We have complex machines and systems as well, where nobody would claim that there are only simple cause effect relationships and you only need a two-page owner's manual.
Drugs work pretty reliably. While you can't predict what exactly is going to happen when you take LSD, you can be pretty sure that something is going to happen. Why, if not because the brain reacting to some chemical?
This idea of "don't say it's a machine, you're oversimplifying it" is what's harmful in my eyes. It leads to anti-scientific positions, mysticism and fatalism. If there's nothing we can do to help, and nothing we can learn because it's so super hyper duper complex and "emergent", why bother. It's not, of course. While we don't know everything, we do know something. And we should extend our knowledge. Pretending that we can't know isn't useful.
In my view its a simple category error, though of a subtle kind. Machines are similar to biological systems in only the most superficial ways: they are non-static systems which exceed a certain complexity and they both follow the laws of physics (see footnote).
There are a lot of ways to think about the differences between machines and people. One very simple one is that machines typically function in only a very limited subset of the space of their available degrees of freedom and small deviations are usually catastrophic. Biological systems don't usually have this property.
Another way to think about it is that machines show evidence of design, which you might characterize as a certain economy of ideas. Two clocks in a computer a much more likely to be identical or to share major conceptual underpinnings than two clocks in a biological system, for example.
I used to have your view of things and then I read Deacon's "Incomplete Nature," a book whose thesis fails to land but is never the less a comprehensive analysis of the properties of living organisms in a unified if rickety system which encompasses the basic laws of physics. It really convinced me that the analogy between brain and machine is poor.
The question of what a mind is and how it relates to the brain is not settled science or philosophy, at any rate. It doesn't do anyone any good to pretend it is.
(footnote): This is a reasonable assumption to make but its hardly widely believed by the general public nor settled scientific fact. Luminaries like Penrose still assert a fundamental disjunction between what brains do and what machines do and he isn't alone among physicists, to say nothing of the broader philosophical community or the lumpen mass of human opinion.
Another thing: to assert that biological systems are just machines is to subscribe to the same error which allows creationists or intelligent design people to put forward the argument that we ought to take the complexity of life as evidence of a designer. It conflates two totally different phenomena by observing a superficial similarity between the two and ignoring all of their manifold and salient differences.
It's just a metaphor. Because the way it works can be viewed as
inputs (sensors, environment, external chemicals) + current state ---> outputs (behavior) + new state
kinda like computers or robots, in a deterministic manner (except it is practically impossible to recreate the exact same inputs, or the exact same state). You don't have to take it too literally, people who read it generally understand what it means.
> inputs (sensors, environment, external chemicals) + current state ---> outputs (behavior) + new state
It's just not a very helpful metaphor, because we are nowhere close to having a way to simulate or compute that function.
So treating people as agents with goals, desires, personalities, experiences and histories and traumas, and drawing on our hard wired abilities for empathy, can be a far more useful model for actually helping and healing people than just giving them a pill.
> So treating people as agents with goals, desires, personalities, experiences and histories and traumas, and drawing on our hard wired abilities for empathy, can be a far more useful model for actually helping and healing people than just giving them a pill.
But that's not a dichotomy. Nobody is saying to either disregard biochemistry and believe that God gave us free will by some magical act or just throw random pills at people.
You can see people as all that and still not reject the idea that chemical imbalances can negatively influence their goals, desires and personalities, and that chemicals can also positively influence them,
This is either very specific and thus not all that applicable to biological systems (that is, the current state is typically much much much more dominant a determinant of future time evolution in biological systems than in machines) or so general as to characterize literally any physical phenomenon whatsoever, and thus hardly useful to think about how biological systems work specifically. Indeed, one element emphasized by Deacon is that the notion of "current state" is ambiguous or at least non-trivial for biological systems and exists for machines only because they represent a major simplification of the relations characterized by your metaphor.
Machines are in some sense constituted by an enormous, intentional, simplification of the possible dynamics of a clump of matter and its relations to the outside world. Their design is profoundly impacted by both our own cognitive limitations and the contexts in which they are manufactured and used. Biological systems simply don't have these constraints and they operate in what are often totally different ways from machines.
My point is that there is considerable scientific and philosophical scholarship on this issue which casts this metaphor into doubt at least with regard to base utility.
I see. Do you think a machine that physically alters its own silicon, or software that rewrites its code instead of storing data, would be closer to biological systems? Or is that part irrelevant?
> This idea of "don't say it's a machine, you're oversimplifying it" is what's harmful in my eyes.
And honestly, there's a lot of complexity to even the smallest thing. A particularly good read, "I, Pencil" goes into how nobody knows how to make even a simple pencil.
I agree, but that seems like a poor example. I could make a pencil out of coal and wood. Wouldn't be as good as a factory one, but good enough. Also, chalk and many other stones/minerals can be used.
Yes and no. The story is more about how that specific (type of) pencil was made, and no single person knows all of it. Even things that we consider simple ... really aren't.
I'm not saying we can't know or that our understanding won't improve. I'm saying right now we pretend to know more than we do. LSD has a reliable effect in the short term, just like torching a plot of land in the Amazon has a reliable effect in the short term. It's the long term effects that are subject to the rules of complex systems and not machines.
But the long term effects are observable as well, aren't they?
I understand where you're coming from, and I'm not too excited when people pipe up with things like "the brain is just like X, so you just need to eat more Y to fix your Z", but we're learning plenty of things about how the brain works, and I believe that we do so because we consider it as a complex system that is not unlike a machine. Remove its fuel, and it will stop operating shortly, cut away parts of it and you'll see functionality change or being lost.
I see no harm in reasoning about it as a very complex and very large machine with tons of inputs, outputs and modification systems. A long time ago, we'd consider the heart a mystery full of magic, and now we're regularly transplanting hearts and I believe we're "within reach" (though rather decades than months) of implanting completely artificial hearts. I see no reason why we shouldn't make progress towards a better understanding of the brain in the same way.
Why? We have complex machines and systems as well, where nobody would claim that there are only simple cause effect relationships and you only need a two-page owner's manual.
Drugs work pretty reliably. While you can't predict what exactly is going to happen when you take LSD, you can be pretty sure that something is going to happen. Why, if not because the brain reacting to some chemical?
This idea of "don't say it's a machine, you're oversimplifying it" is what's harmful in my eyes. It leads to anti-scientific positions, mysticism and fatalism. If there's nothing we can do to help, and nothing we can learn because it's so super hyper duper complex and "emergent", why bother. It's not, of course. While we don't know everything, we do know something. And we should extend our knowledge. Pretending that we can't know isn't useful.