Prediction: If posting = mining, then by 2020-2025 AI posting and curation will become the ASIC-equivalent of mining

Since this is a blockchain-based project, and things will remain in the blockchain forever, I want to make a time-stamped and blockchained prediction for the future of Steem (or Steem clones).

Given that some posts in Steem actually generate quite some money - for something like "blogging", I think it's a matter of time before some people start "dissecting" the "recipe of success" in a ...scientific way.

What I mean by that is that statistics will start showing patterns emerging. Patterns related to things like length of post (pages), number of images, bullets/numbering, number of words, time required for the average user to read through it, trends regarding use of caps/small/first letter capping, trends regarding proper use of language, trends regarding the attractiveness and gender of the author, keywords and subject matter, length of title, wording of title, coloring and content of images used for the main listing, time posted, time posted relevant to other posts posted in the same time period which may be "junk" (thus elevating perceived value) etc.

The problem is that humans are crappy at pattern recognition. This means that their analysis will always be sub-standard. So... As futuristic as it may sound (self-driving cars weren't futuristic?) I estimate that by 2025, Artificial Intelligence software will be able to actually create quality content. This will be a result of fusion with AI-pattern recognition abilities which find successful trends in content. Thus AI will be able to post proper content at the proper times to maximize its income and compete (!) with human authors. Initially it will probably start with AI trying to insert small comments that have high-upvote possibility and then escalate to meme-type posts, before creating full-blown posts itself.

I also expect that by 2020 (latest) AI curation will take the place of if-then-else bots - based on finding common patterns of success in curation / rewards.

If posting and curation is the "mining" of STEEM, then I predict that AI will be the "ASIC miner" equivalent. 

In the meantime, and until AI of such complexity is developed, humans will probably have to do the pattern recognition analysis themselves to find those patterns which are significant for success both in posting and curation.

Let's see how the prediction goes...