Machines Before The Machines
03 Jun 2016This is an argument for biology as a viable medium for engineering.
Biology Is Soft
Biology has failed to find a central/ general/ formal framework for thinking about organisms. In this way it’s always been unlike its peers, physics and chemistry, which translate much better to the applied domain.
Sure, we have the central dogma: DNA \( \to \) RNA \( \to \) Protein. Which gives a high-level picture of how biological systems work, and implies how they’ve come to be in the first place (the primordial soup).
But if we’re being honest, it is really just a sketch. It’s not formalized and it’s not general. If we want to make biology harder, so its applied domains have a platform to really stand on, it must have some more formality and generality.
Another example, the theory of evolution. This is hands-down the central paradigm of biology (making the dogma a close second). Simply because “nothing in biology (including our understanding of the central dogma) makes sense, except in light of evolution.”
And yet, the theory is soft, too. Theories are “hard” insofar as we can use them for making/ designing things, and we can’t really do that with evolution. Sure, we can use it in a vague way to design/ build systems that have the capacity for evolution and so on; but this is imprecise business. Where are the numbers?
Bottom line, biology isn’t there yet, it’s still soft. I believe undoubtedly the answer has to be mathematics and computation, as they are the most effective tools we have.1
Machines Before The Machines
The naysayers whom doubt that biology can become as hard as its more formal counterparts may be right. Chemistry is less formal than physics so it seems reasonable that biology would be less formal than chemistry. This may be unavoidable.
But the way I see it, biology, the study of biological organisms, is much like the study of automobiles—their varieties/ styles, their organization, their evolution. (Yes, machines, organisms are like machines.2) So, in theory, biology should at least be as formal as a study of mechanical/ electrical engineering (which are less formal than physics by virtue of being applied).
Now by this point, we’ve examined them enough to understand (vaguely) how they work/ operate at a low level. And also for the big bulky, highly mechanical ones (which we call mammals) how sometimes to fix them when they’re broken, at a high level. But we still haven’t understood them to the point where we can design and build them at all levels.3
And we should remember this is possible. Because, in truth, biological systems were/ are the original machines. They were the machines before there were the kinds of machines we think of as machines (automobiles, robots, etc.), the kinds we design and build at all levels.
Towards Value
And furthermore we should remember that the power/ value/ usefulness of a machine is that it does work. It transforms some (usually more available) energy/ material (e.g. gasoline) into some (usually less available) work/ energy (e.g. locomotion).
So with respect to biological systems it shouldn’t just be about making small genetic changes (although this is inevitably where we must begin). But in using the theories of biology (again, in more formal terms) to engineer biological systems that trully harness the power of their machinic nature (the production of value).
1. It’s been argued that the most viable candidate for a mathematics suitable for the formalization of biology is (so-called) post-modern mathematics. I happen to agree. I think that theoretical computer science (and related maths) has much to offer biology in the way of formality. Theories of trees, learning and language; it seems like it’s all there, minus maybe some statistical mechanics.
2. Except, for one, biological machines are made of biochemical components, which makes them much more complicated and dynamic at their lowest levels. Mechanical components, on the other hand, are much less dynamic, more homogenous and do not need an aqueous environment to operate. And, for two, they’re machines we didn’t design, only operate, so we don’t fully understand their internals, especially how they work between levels. (We are exactly like most operators of modern technology. We don’t have to know how the internals of our car/ smart-phone/ tablet to work the user interface.)
3. At least now we have the right idea: genome engineering initiatives (Human Genome Project-build), the rise of CRISPR and synthetic biology research/ industry will no doubt increase our toolset. Systems biology research hopefully will increase our understanding of the underlying dynamics.