http://assets.rollingstone.com/assets/2016/article/inside-the-artificial-intelligence-revolution-a-special-report-pt-2-20160309/231324/large_rect/1457449175/1401x788-AI-opener.jpg
I already see something happening on Steemit that i've not generally seen on social networks - international collaboration.

## A subtle but critical difference

With Reddit, Twitter, Facebook, the amount of "user clustering" is actually quite high.  Facebook, it is because users generally are meant to "know" each other in real life (often false, but this is the core) and if not, then users are generally in the same geographic area.  

With twitter, the low word count limit causes heavy ideological clustering.  If all you can write is a bumper sticker, you end up with a bunch of yes-men types.  I can't stand twitter for this very reason.  Lots of trite nothings forever.

Reddit is organized around topic - which is a different approach.  However, each post is generally only in ONE topic.  Steemit is interesting in that it allows the same post to be in 5 different categories at the same time.  This is an interesting property because it means something in chinese can still be posted to tech and be read.  As machine translation gets better and better, this will tear down barriers.  These extra categories will really help translation as well.  If the machine knows you are talking about technology, it can do a better job with translation.  For instance, in vietnam the word for water "Nuoc" and country "Nuoc" are the same.  They are just used in different contexts and in different ways.  An article tagged "politics" will help the machine decide if country or water is a more likely fit.

## AI will unlock the power of blockchain networks
http://www.alvervalleysoftware.com/wp-content/uploads/2015/01/dreamstime_xs_29076006-edit.jpg
AI is already being used to write articles both here and in newspapers.  While I worry that this will devalue human content, it also has the potential to unleash it.  If AI were tuned from content curation and creation to FEEDBACK then that would  help humans get better and understanding their own communications.  Right now tons of good content doesn't get rewarded, and sometimes mediocre content does.  This is because of how things are built - popularity can be objectively measured, while supposedly quality "cannot".

## Metacuration is the answer
I don't think this is true.  What is popular is not always good (mostly it is what is most accessible) but just because popularity is easy to measure doesn't mean quality cannot be.  In particular, If AI started seeing the connections between what users were interested in and directing users towards those posts - that would be a curation FAR superior to just upvoting.  And if AI used the same connections  to help authors create content that better helped their readers then that, too, will be a far better thing (and objectively verifiable) than mere popularity.

## The best uses of technology are not to judge, but to enhance
I often worry that technology in general is doing less and less to enhance humanity and more and more to replace and/or extract value like an oil well - pull and pull until it's empty and then leave.

But it doesn't have to be this way and I think steemit can evolve to show the way forward.  It's already doing a better job than the others (whatever the rough edges), and with good guidance it can do even better.

## <center> Want more art, philosophy, and travel?  [Follow @rampant!](https://steemit.com/@rampant) </center>