This doesn’t work, and if it did, it wouldn’t be evolution. The latter first: For AS, or GE, or ATI/ATT, or epigenetics, etc., to create a human from a primitive ape would mean evolution had to have created 25000 genes, and the molecular mechanisms to create that change. If a factory created a bunch of parts which just happened to fit together to make a car, would you say it evolved? Of course not, this is silly. The fact that you insist this is evolution suggests you want evolution to be true. Indeed, even you had to admit how remarkable this is:
Indeed, from an evolutionary perspective this is astonishing. What luck? The level of serendipity is astronomical. Can you imagine the chance evolution just happened to create all that stuff, which incredibly turned out to make humans? I guess you could appeal to the anthropic principle, multiverse, and so forth. Evolution just keeps on surprising us.
Now the other problem is that all of this is unlikely anyway. The vast majority of our genes are alternatively spliced, and science indicates that, surprise, this is not merely useless diversity or junk DNA doing its thing. As this paper explains:
The history of evolutionary thought is full of claims of uselessness, only later to proven wrong. Alternative splicing looks to be quite useful and important. We have to assume that most of the human alternative splicing is important. No, that’s not to say you can’t change the splicing anywhere. But you probably can’t just have any old splicing patterns. The different alternative splicing of a gene can result in a very different gene product. So it is not good enough to just generate tons of AS, willy-nilly. It has to be the right AS, out of an enormous space of possibilities. Secondly, gene products typically work together. Sure there are singletons, but often proteins form quaternary structures, pathways etc. So most of these AS genes are going to function, at least part of the time, in a network, they need to come in groups. So a trillion mutations over 6 MY is not going to help. Drift over MYs is going to give some winners, but mostly a whole bunch of incoherent combinations. Selection isn’t going to work because you can’t get a group of newly created, coherent, ASs in one individual. You’re going to need a whole bunch of multiverses for this.
It sounds so easy doesn’t it? Evolution creates the genes, the introns / exons, the splicing mechanisms, the TFs, binding sites, and all the other regulatory machinery, and then it proceeds to experiment around, mutating these and building up a huge library of neutral functionality, and then at some later time, given a critical mass, or the right environmental shifts, that functionality now suddenly provides a coordinated orchestra of organization and emergent functionality. This is how evolutionists view the biological world, without concern for the serendipity.