the fifth miracle , page 66:
In a living organism we see the power of software, or information processing, refined to an incredible degree. Cells are not hard-wired, like kites. Rather, the information flow couples the chalk of nucleic acids to the cheese of proteins using the genetic code. Stored energy is then released and forces are harnessed to carry out the programmed instructions, as with the radio-controlled plane. Viewed this way, the problem of the origin of life reduces to one of understanding how encoded software emerged spontaneously from hardware. How did it happen? How did nature “go digital”? We are dealing here not with a simple matter of refinement and adaptation, an amplification of complexity, or even the husbanding of information, but a fundamental change of concept. It is like trying to explain how a kite can evolve into a radio-controlled aircraft. Can the laws of nature as we presently comprehend them account for such a transition? I do not believe they can. To see why not, it is necessary to dig a bit deeper into the informational character of life.
A living cell is made largely of proteins. This is the hardware. The membrane surrounding the cell is analogous to the plastic shell of my computer, or, perhaps more accurately, to the microchip substrate onto which the circuitry is etched. It’s no good, however, just throwing a heap of proteins into a container and expecting life to happen. Even with the necessary raw materials, a cell won’t do anything clever without software. This is normally provided by DNA. Like the floppy disk, DNA is itself hardware, but again the crucial feature is not the stuff of which DNA is made but the message written into its base pairs. Put this message into the right molecular environment—in the right semantic context—and, what do you know, life happens! So life is a felicitous blend of hardware and software. More than mere complexity, it is informed or instructed complexity. Let me illustrate this subtle but absolutely crucial point with a couple of analogies. The nineteenth century was the great Age of the Machine. Many clever devices were invented. Take, for example, the steam-engine governor, a pair of balls attached to levers that rotate at a rate determined by the steam pressure. If the pressure gets too high, the balls whirl so fast that, by centrifugal force, they lever a valve open, thereby reducing the pressure. Today we would describe the principle behind this type of mechanism as “feedback.” You wouldn’t do it with balls any more. Instead, a sensor would feed data about the pressure electrically to a small computer or microprocessor. This electronic system would then process the information and instruct the valve to open or close using a motor. My wife’s Holden Berina car has one of these microprocessors to maximize fuel efficiency. It decides how fast the engine should run when it is idling. The difference between the push-pull mechanical steam governor and the electronic microprocessor is that the former is a hardware solution to a problem and the latter depends on information processing and software, i.e., it is “digital.”
The power of software is that it can act as an interface between chalk and cheese—different sorts of hardware that otherwise could not deal with each other effectively. Compare the difficulty of trying to steer a kite with the ease of flying a model aircraft by remote control. The difference here reduces to hardware versus software. The pull of the kite strings is a direct but very clumsy way of coupling the kite hardware to the control hardware (the person on the ground). The radio system, which first encodes the instructions and then relays the coded data to be interpreted at the other end, works much more efficiently. Of course, the informational flow from ground to aircraft may also be described in hardware terms: radio waves propagate from the transmitter to the receiver, where they induce an electric current that triggers circuits and moves airfoils, etc. However, this hardware description is merely incidental to the performance of the plane. The role of the radio waves is simply to serve as an information channel. The waves themselves don’t push and pull the aircraft about. Instead, the coded information harnesses other, more powerful, forces to do the job. A lumbering kite is a (literally) hard-wired mechanism, whereas the more efficient radiocontrolled plane is an information-controlled mechanism.