That may be true, however, the DNA 'dice' are not loaded. Chemistry does not force in any way a particular order of the DNA molecules. There are two proofs for this:
1. DNA molecules do not touch each other in the linear sequence. Instead they are joined 'on the outside' by sugar/phosphate molecules which make equally easy bonds with all four different DNA molecules.
2. If there was a chemical necessity to any particular scheme, we would be seeing that certain possible combinations do not occur. Instead we see all 64 possible combinations of the three bit code appearing in living things.
What you are saying is not strictly true, but the point is minor enough that it doesn't really change things one way or the other.
More importantly, DNA doesn't "do" anything, merely providing a template. Building proteins off that template is an extremely biased system, and we expend a fair portion of our supercomputing power today figuring out what protein conformations are probable under certain circumstances and which aren't. The combinatorial space IS astronomically huge, and sifting through the vast quantities of chemically improbable combinations to find the small number of highly probable combinations is not trivial and won't be for a long time, running into what is essentially an expression of the Halting Problem of computational theory.