Brief idea about Steemit, monkeys on a keyboard, post005
I was having a conversation with @somethingsubtle here: https://steemit.com/steemit/@somethingsubtle/a-tale-of-two-steemits#@spbeckman/re-somethingsubtle-re-spbeckman-re-somethingsubtle-re-spbeckman-re-somethingsubtle-a-tale-of-two-steemits-20180520t042804550z
and I think I came across an important idea. I'm going to refine this some later, but for now, I want to simply get it posted to discuss it with others.
My comment to @somethingsubtle was:
I don't know if I'm an optimist, but I'm trying to understand what Steemit is about. As it is, people just seem to post lots of random "stuff" most of which is low quality. If this is all that Steemit is, then it might as well not exist. (I don't like reading most of what is on this site.) However, if we pay for curation, possibly collection, and at some point start hiding the garbage, we could have some very nice content.
I'm imagining Steemit kind of like the room full of monkeys and a typewriter that given enough time will produce Shakespeare.
In the case of Steemit, we (sort of) have a means to train the monkeys. We also have slightly more talented monkeys that are trained to curate content. Lastly, we have developers and possibly machine learning tools, that will develop algorithms to clean up some of the mess left by letting the monkeys bang on the typewriter.
Is that optimistic? Maybe? Perhaps delusional. :)
I should mention that my research is in the use of machine learning and informatics methods to study materials science and engineering. I take low quality, busy, noisy, messy data and use algorithms to find order and understanding. From my perspective, there are likely some real gems that can be refined from this mess of content.
What I've realized, is that this is the future.
There is no way to fairly censor the contribution to say that, "Unless you're educated enough, or connected enough, you're not allowed to contribute." That model is gone, and we're better off for it. However, as a result we have places such as Steemit that are full of low quality content. Most of what's posted here the world would be better off without. (For heavens sake, there are people whose entire existence is about writing posts to encourage others to post.)
How do we find content in the noise? Machine learning and data informatics. I don't know exactly how this is going to work yet, I think that we need algorithms that have a better understanding of language, but ultimately, we'll need some algorithm to extract the creative and original ideas from the noise.
To explain a little, in my research mostly I have noise, but the algorithms are able to find the hidden patterns and the underlying meaning. When I was learning these methods I used the grade book for the course I was teaching. I messed up the data and filled it with noise, but the algorithms were still able to accurately classify the students according to the grade they should receive.
So from my perspective all the superficial stuff that happens here, the plagiarized content, the fluff posts, the upvote bot posts, is not a problem. OK, it is a problem in the short term, but in the long term it isn't. Most of this is going to go away and be refined. The Steemit users, and curators, and admins are data producers -- some of it is useful, most of it isn't, but from this steeming pile of data we'll produce gems.
This is just a short post to stimulate conversation. Thoughts?
Consider me stimulated.
Ultimately, I think you are correct. The spam on Steemit will slowly die away as algorithms improve just like spam emails. Spam emails are still created, and they are still sent, but they are no longer seen; therefore, they are no longer a problem.
Steemit should eventually evolve that direction where low-grade content is more easily scrubbed or hidden from the platform completely.
If the platform survives this ugly and painful birth, it has the potential to really thrive.
That's how our bodies work... Often chaotic and seemingly broken when you get in and take a look at the cellular level, but functioning effectively as a holistic system, nonetheless.
You can think of complex systems as being build of smaller, very simple, irreducible parts. When there are enough small parts, that all interact together, we have what's called an emergent phenomena. The "Game of Life" was one example of this. You might find this neat:
https://en.wikipedia.org/wiki/Conway%27s_Game_of_LifeIn terms of the monkeys I'm thinking of them almost as automata that make random acts on the keyboard. Then we need a way to look at their acts and discard the ones that aren't productive toward producing Shakespeare and keep those that do. Something to shape their collective motion.
Hi @spbeckman
My reply got a little long so I decided to post it as a blog to see if I could pick up some cigarette money. Here is the blog:
https://steemit.com/blockchain/@transhuman/my-reply-to-spbeckman-regarding-his-article-titled-brief-idea-about-steemit-monkeys-on-a-keyboard-post005
Your ML expertise is interesting because I am interested in seeing AI digest the scientific literature to find a cure for aging. I own a Google+ Community where I post about the topic here: https://plus.google.com/u/0/communities/101513179882783511389 (Google+ account required for full access).
I see it, but haven't had time to think about it all yet. I have a proposal going in tomorrow and have spent the past 48 hours pretty much focused on this. I'll write something later in the week.
Yes and no, I think...
My understanding if that blockchain isn't about crypto calculations, but it is simply a means to record that something was done, and the record can't be corrupted. The first application of blockchain was Bitcoin and the work done was the passing of coins from one person to another.
What initially bothered me about Steem was that I didn't understand why they didn't need miners and all that overhead found in bitcoin. The answer is that they simply make new coins every day without work and they record them in the blockchain.
I could be completely wrong about this... but as I said, this is the understanding that I've come to the past several weeks.
So... one thing we could do is have blockchain record peoples contribution to high performance computing. This is something that Gridcoin is doing.
https://www.gridcoin.us/Another thing that we could do is simply use blockchain to record data. For example, after the presidential election I heard about climate scientist backing up data to blockchain. (I have no evidence that they actually did this, but such is the nature of rumor.)
I'm not thinking about blockchain in these fashions, but more in terms of using the social aspect. I imagine it as a means to create multilevel, hierarchical social structure that rewards human work. For example, I'm interested in a particular metal alloy that has been studied for just short of 100 years. There is lots of data, that is stored in print. It is being digitized, but just pdf scans are being made. Can we pay using Steem, or something similar, to have people strip out data, say microscopy images, and record for each the details, such as magnification, contrast, composition, etc. Then can we pay curators to verify the work. Then pay a second level of curators to organize the data in a meaningful fashion? Then pay a third level of curators to.... etc.
I just submitted a proposal to create a database and as part of this we had to design a means to encourage people to contribute their data. (You may not believe it, but scientists talk about data sharing, but actually don't do so.) The whole writing period I was reimagining it built on steem.
We'll see what comes from this. Before I get ahead of myself I need to get some science content up here to build my credibility with the community. I also would like to figure out how to make a buck on this system. From what I can tell, my style and time for posting (early mornings and on weekends) make it likely my payouts will be low.
This statement "pulls the chain" of a many a Steemer and other cryto-nerds. The watchword in crypto (blockchain etc.) is: Decentralization ; disintermediation via managing systems over peer to peer computer networks. For example, they want to deprecate banks and fiat currencies. You can send a bitcoin to another entity without going through a bank. Bitcoin goes from the wallet of one person to another directly without bankers interfering.
Hello, @spbeckman I just checked your introduction post and I know u would be in love with the steemstem community, follow us on discord share quality content and get curation rewards up to $120 . Especially as a professor in engineering I know you will thrive in the steemstem community and make the community a better place witn your knowledge.
See you there!
Steem your way to power greatness Nd glory !
I just signed in... but I don't exactly get it yet. Is there documentation? It is very non-intuitive to me.