If you share my view that technology drives history more than any other factor, then you will probably agree that the 21st century is going to be significantly shaped by the outcome of a single question: Will synthetic biology achieve radical success or not? In this column I’ll describe an early warning sign to watch for that will give us a clue about which way this important new field is headed.
Synthetic biology is the current term for the outer reaches of ambition in biotechnology. More often than not, the notion includes making artificial biology more like digital computation. It could hardly be otherwise, for computers are central to most of the prior art we have for building highly complicated structures from scratch. Computers also symbolize the ultimate in freedom through technology. You can hypothetically program a computer to do virtually anything with its input and output devices. If we could only find the right computer program to operate robotic medical devices, for instance, we could create a robot surgeon to cure any disease. If we could do the same with DNA and the other chemicals of life, we could create a huge variety of novel creatures or transform ourselves into astonishing new forms.
But if we entertain the idea that biotechnology is going to become more like computation, we aren’t being very specific, because there is more than one kind of computation. In particular, it might be more revealing to ask if synthetic biology is more likely to turn out like digital hardware or software. That’s an excellent candidate to be the most important question of the century.
From a mathematician’s point of view, hardware and software are practically interchangeable. You can almost always emulate a chip in software or implement a program as a chip. In practice, though, the two things could hardly be more different. Chips get faster and cheaper at a predictable, accelerating rate that is so reliable it is known as a law—the famous Moore’s law. Software typically gets worse over time.
It’s true that faster computers enable new software algorithms that weren’t possible before, like ones for machine vision (see Jaron’s World: Computer Evolution), but old programs don’t necessarily get better as hardware improves. In fact, they often lose efficiency at such a breathtaking rate that they effectively cancel out Moore’s law when they are adapted to run on new, faster machines. Try opening a similar word-processor document on old and new computers: The performance is often similar, even if the hardware has improved a thousandfold. How can this be? Software is so difficult to work with that in practice it almost never achieves its theoretical potential.
If synthetic biology turns out to improve in the accelerating way that computer hardware does, we will be in for quite a ride. It’s hard to predict how weird things could get, so one is tempted to max out deliriously as a futurist. Imagine an artfully designed fungus that looks like a hat; when you put it on, it digests your head and turns it into a still-conscious, rubbery Super Ball an inch across, suitable for easy launch into space. Once there, another fungus might then reconstitute your head and form a protective life-sustaining bubble around it. (This prediction may go too far, but the point is that it’s hard to say by what margin.)
If synthetic biology instead turns out to be more like software, it will still be amazing but in a more incremental, less predictable way. We will witness a succession of plateaus of achievement in areas like medicine and bioenergy. After a decade or two, we might have engineered bacteria that make fuel out of old garbage dumps, or maybe even a substantially artificial cell that acts like a doctor, swimming through the body and fixing our own aging human cells.
Then again, reality often violates our preconceived notions, and synthetic biology could turn out to have a character that doesn’t resemble hardware or software. Natural biology is certainly unlike either of those! It is flexible, as software ought to be from a naive point of view, but it is not as fragile as software. Synthetic biology may very well introduce a fourth kind of design complexity that has some of the qualities of all three precedents.
