Genes produce proteins that construct the body.
It's a simple statement, yet incredibly deep. We know that it's true because we can observe the process, yet we have almost no understanding of how the process actually works.
I propose to try to build a machine that will give us this understanding.
The machine is a simulator, written in software. It is a simulator that executes a particular program, running it from start to finish and allowing us to see the results at any stage.
The program is a genome written using the language of DNA. What the machine does is to take a genome, and execute it.
Let's start with some of the simplest lifeforms we know about, and try to "grow" them in the simulator. We have mapped the DNA of many simple creatures. We know what the DNA is supposed to grow into. So we have a control that lets us ensure the machine works.
It's a hideously ambitious project. Simulating the chemical processes that happen in a single cell is complex enough to be beyond the reach of today's technology. We can simulate a virus and some other simple forms of life. But a cell is much, much more. Simulating an organism that has trillions of cells is so complex that it boggles the mind.
And yet, technology will eventually let us do this. The question will not be one of power and capacity. It will be, firstly, whether it's possible at all, and secondly what such a machine could be used for.
The first question is a bit subtle. Take a closer look at life and you see that DNA is not like a CD-ROM, an isolated piece of software that you can run at will. DNA lives in a sea of life, and all life is interconnected in many ways. Our reproduction and growth depends on the DNA of other organisms: bacteria, mainly. It depends on many other mechanisms based on RNA. Life depends on the unbroken chain stretching back to the very first ancestral molecules.
I suspect that no modern genome can be executed without bootstrapping of some kind. A set of chromosomes is not sufficiently complete: it requires the bootstrapping provided by the cell. Perhaps the very first lifeforms, consisting of RNA, could be simulated with no cheating. But modern DNA needs proteins and modern proteins need DNA.
It's quite possible that there is no way to jump into the middle of this running program. Certainly it should be feasible to "snapshot" a lifeform, and then continue to play that snapshot. But I suspect that it is impossible to snapshot a realistically complex organism. I'd go fifty-fifty on this.
As for the second question, assuming we can build the machine, what could we do with it? Imagine being able to take a hair from your head, scan it into a computer, and watch as you grow from a single cell into a fully-formed person. You watch yourself approach your own age, complete with fashionable tattoes and piercings. You watch yourself grow old, wrinkled, and finally the clock stops and says "DIES OF CARDIAC ARREST AT AGE 62.3". Possibly a little more information than you wanted to know.
If and when we can build the machine, we will be able to create life, literally. It is a small step from scanning DNA to constructing it. A DNA printer could take a genome and encode it into a cell capable of growth and reproduction.
Imagine being able to create lifeforms. Take a little chicken DNA, splice it with penguin DNA, and you get a bird that swims _and_ tastes real good.
Unlimited freedom to manipulate genes sounds like a nightmare scenario for most people. But like all technology it would be largely self-balancing. For every terrorist constructing the ultimate virus, there would be someone building better immune systems.
I predict three things.
First, that such a machine will be built and it will work. Given the scale of the project, I guess it will be at least 50 years from now.
Public attitudes towards such manipulations will soften as we come to see life as a sea of genes, rather than a hierarchy of organisms. Religious fundamentalists may try to ban such science, but the machine will be built, and used.
Secondly, the technology will follow the standard technology curve for immaterial products: it will become cheaper and cheaper until within two generations, it will be accessible to school children.
So the timespan is a century or so.
Thirdly, this technology will change life as we know it. Today's so-called "genetic engineering" will pale in comparison to what our great-great grand children will do with such tools. They will construct new lifeforms, for pleasure and business. They will change themselves and their environment beyond all recognition. Finally they will take control of the economy of competition and collaboration between genes that is the basis for life itself.
We are what we are - good and bad, heaven and hell - because our genes make us so. Just how a DNA sequence producing proteins can turn a single cell into a homicidal maniac or a gifted writer is a mystery to us. But it will be basic science to our descendants. What they do with that is anyone's guess.