Could Big Publishing put up a website to compete with Amazon's?

Someone on KBoards started a thread about whether readers would flee to the Big Six (actually the Big Five now we have Random Penguin) for guidance as to what to read, faced with a flood of indie offerings of mixed merit. Perhaps they might band together and create their own website to rival Amazon, selling only quality books from approved writers? KB Burke's comment was so interesting I asked his permission to quote it here:

Just some observations from a software developer ...

The big publishers haven't figured out discoverability. Their talent pool means little when readers can't find the books. Just because they put 1,000 vetted writers on a site, doesn't mean I can find the 12 that I want to read.

To do that, they need years of buying behavior. The science behind 'also boughts' is called Collective Intelligence, and making the website is 10% of the battle. Basically, you model the mass behavior of people, and then identify patterns. 16 year old guys who bought Warhammer books, also bought Halo books. If you're 16, and have bought one of these things, you might like the other. And so on. It's statistics, but you need data to build the model.

This is why Amazon bought Goodreads, for the data. If the Big Six understood software engineering they would have bought that network years ago. Websites don't sell books. Data sells books, and the big publishers don't have the data to compete. They are 10 years late to the party.

This why Amazon is always tweaking their algorithms. They get more data and adjust their models. It's no different than a presidential campaign modeling an election by 'likely voters.'

Dozens of tech companies, with big time talent, like Apple and Google and Sony, have failed to compete with Amazon. They don't have data, but they do have some of the most talented engineers in the industry. Think about that. No one in New York will have anything like Google's resources, and Google isn't hurting Amazon at all.

As far as quality goes ...

There might be 100,000 bad indie titles, with quality issues, but it is probably a bell curve. Some percentage are high quality, 5-10%, that compete with the big publishers.

This has been true since the 1930s and the beginning of pulp. There is an ocean of crap, and a small handful of standouts. Who curates that crap doesn't matter. This is why word of mouth sells books. Amazon has made significant leaps in this regard, with their algorithms, but no one else is close.

The ocean is bigger today, but the model is the same as Edgar Rice Burrows. His books sold, despite that ocean of crap.

Collective Intelligence is really why Amazon dominates the book industry. The traditional publishers have a team of editors telling me they found another Edgar Rice Burrows. They are telling me what I should read. Meanwhile Amazon is telling me what people do read. People who like Burrows have also bought x, y and z. This helps me find my tribe, so to speak.

This is the now. The automation of white collar jobs, like book curation, and it won't ever go away. New York thinks good taste can't be automated, but that's because they don't understand the science.