Because until we get mind uploading, there aren't enough smart people to cover everything (e.g. even though Tao knows way more math than the average mathematician, I have solved interesting math problems that are simply not in his or in any other genius's immediate interests).

Why do anything then? It's a very nihilistic attitude and a very narrow conception of what "smart" is. How do you even begin to compare? How can you say that one author is smarter than another and why use some abstract metric that attempts to measure a characteristic of the author and not the works themselves? How do you not consider the audience when assessing the quality of the writing? And then there will be just personal preference.

Gist:

Writing helps to truly understand something, and to think better

Your experience and context make your writing unique

It matters little that they’re smarter people out there as long as it is basically random chance that determines whether a reader discover them or those who are dumber. A way to find the absolute objective best in any field - experts, products, methods - would be revolutionary.

As of now, the best thing we have is marketing, and politics. Not he who is the best is most easily heard, but he who screams the loudest.

Because knowledge is a web of associations, conversation (and writing) is an imperfect means of conveying those associations, and there is always a chance that any new explanation of something might convey some connection that was previously elusive.

There may be others who are smarter, but that doesn't mean they write better, or even bother to write at all.

Comparative advantage: your opportunity cost is lower.

The smart to stupid is a smooth curve, so everyone have a point in the line.

Writing is a way for you to discover what's smart about you