Bottom Up Design, the Second Creative Force
There are two methods by which things are created or organized. The first method we all understand because we do it every day. We think, have a plan, then we move things around and make it happen. This is top down design, it starts with a designer.
But there is a second method that is often overlooked. A bottom up approach, where from tiny parts and simple rules, a design emerges. The end result can look as if it was created. A good example would crystals. For example, how does a quartz crystal form and get its characteristic look?
A quartz molecule looks somewhat like a pyramid with magnets on all the far points. These click together and form shapes, because any molecule out of place is repelled and any molecule that sticks out might snap off. Also new molecules attach more easily to incomplete sections than a smooth surface. This grows the crystal in layers. But the larger a layer is, the longer it takes to complete. So the small triangles on the top complete faster, growing the crystal in length more than in width.
This kind of emerging shapes are everywhere. Pull the plug from your bathtub, and a twister appears. The same twister appeared in space when gravity sucked up matter to form the sun. The twister flattened and formed the planets.
Birds flying in flocks need no conductor, instead, every bird just keeps to three simple rules: stay close to other birds, don't crash into other birds, follow the general direction of other birds. Just those rules create a murmuration of birds. Or a school of fish. No top down coordinator, but simple rules for every bottom up entity. And the self organization emerges.
A big example is life. Its rule is extremely simple: make copies. From there two effects emerge. First there is a doubling effect. It starts off with one, it makes a copy, now there are two. Both make a copy now there are four. Doubling for ten steps and there are a thousand. Another ten steps give a million, another ten steps give a billion.
The second effect is self improvement. If not every copy is perfect, some copies might be broken and don't copy. Some copies might be bad and copy slowly. Some copies might be better and copy faster. The broken copies do nothing. The bad copies grow slowly and do little. But a faster version will outgrow and replace the original version given enough steps.
Will a faster version appear? Around step 30, a billion copies have been made. If one in ten is not perfect, there are a hundred million imperfect variations. Every variation is an experiment to find a faster version.
Moreover, versions where every copy is a variation will not have a doubling effect. Versions for which almost every copy is perfect will grow quickly but do very few experiments. Versions that strike a good balance between copies and variations will find faster versions the fastest.
From the simple rule of self copying, emerges the fastest copier that experiments the most.
But there are limits. Doubling will slow down as building materials run out. Copiers might break down, probably becoming building materials. And perhaps some copiers start using by-products of other versions, or even start using other versions wholesale as building blocks. Complex interaction might appear. The fastest copier might not be so easy to identify. Perhaps allowing for many good copiers, each in their own category.
Life we see around us is self copying. Life's chemistry supports it. A question is what would be the simplest self copying system or molecule that the chemistry supports. And could life as we see it today emerge from such extreme simple beginnings. Some of this is still an open question. But life has all the hallmarks of bottom up design.