RE: Attention-Based Stigmergic Distributed Collaborative Organizations
@dana-edwards your post and topic are insightful. But pseudo AI networks based on human attention are not new. We've had them for a long while now.
Your idea while innovative is not unique, both Amazon Mechanical Turk and Clickworker do exactly as you describe.
There isn't any benefit that I can see to combining this with blockchain tech the way you described. Nevertheless it is really good food for thought and I would like to collaborate with you on a real distributed AI targeting CDSS applications, but one supported by human experts such as licensed clinicians especially during it's earliest training phases.
Our plan is to make this an XPrize submission.
Let me know if you're interested and I'll contact you offlist with your permission.
p.s. While I realize your primary thrust is humans, you may be interested in reading this...
http://www.wildml.com/2016/01/attention-and-memory-in-deep-learning-and-nlp/
What I'm talking about is intelligent agents. More like Amazon Echo than Amazon Turk. More like Alexa than Clickworker. Here is an example of what I mean: https://github.com/cantino/huginn/
It is my understanding our brains are the bottleneck. Relying on human experts makes sense when there are experts but when you're doing certain new things there is no such thing as an expert. At the same time, how do you know which people you can trust? So instead I'm thinking intelligence amplification is the better solution and the reason to do it on a blockchain is so the whole world can have access without a point of failure.
In your post on your blog you mention beneficial bots. I'm very much for beneficial bots with the understanding that anything which saves time or attention adds value. Both time and attention are scarce resources for humans. If we can set curation on autopilot that is fine as long as it's voting based on criteria we would vote on if we had the attention.
Hey Huginn looks like a really neat platform. Thanks for showing that to me, I can already see possibilities.
I have to disagree on the brain being a bottle neck. It's by far and away the most powerful parallel processing system in the world. However I do understand your point. There are only so many hours in a day and it makes sense to automate repetitive tasks. It's the body that's the bottle neck though, not the brain. Just think about how we develop workflows in our daily lives that begin to self optimize and I think you'll see I'm right there.
We'd have to ask @dan to be certain, but I'm reasonably certain that Graphene, the blockchain tech that powers both bitshares and steem would never be able to deal with the kinds of irregular data you would need in order to store any sort of bespoke knowledge.
Blockchains are not a particularly well suited storage mechanism for this sort of information though because they are really nothing more than a distributed timestamped ledger service with cryptographic guarantees. If you don't need a rigid form of consensus then you don't need a blockchain. What you describe is the opposite of rigid consensus, it's fluid consensus based on perception. "I 25% agree with person A and 40% disagree with person B" In the blockchain world that's a consensus fork. In the real world, it's just how people are.
There is however the possibility of combining cryptographic signing and authentication with a regular document store or even a graph db such as neo4j. So long as there doesn't need to be rigid consensus, just regularity of the dataset then this approach makes sense. For distributed raw data storage you could just use Freenet or something like it...
https://freenetproject.org/
On the whole I do agree with you. Intelligence amplification as a force multiplier is the way forward. And the real secrets will be found in small specialized agent systems that can act cooperatively, i.e. swarms.
If I can be free to think, I can think more. Even typing is a chore that would better be handled by dictation AI. Curation based on a rudimentary form of collective swarm intelligence but with human guidance is also the topic of the blog post I linked above and really, really hope you'll read and comment on. :D
I'm not sure if Graphene in specific could do it as it is now but I do think it could be upgraded to manage it. I know for sure Tauchain could do it which is why I've given so much of my attention to Tauchain. Freenet is a bit outdated (2001) technology and doesn't even handle basics of the semantic web so I would not rely on it. The technology you need is the technology behind the semantic web which Tauchain promises to have. Solid also looks good and is made by Tim Berners-Lee, designed specifically for social media.
Generally though, it's going to happen sooner or later whether we see it on Tauchain or Graphene. To me a blockchain is just a data structure. But it does not have to be rigid, you can make it flexible and Tauchain shows that, but even to some degree you see it in DPOS. The blockchain allows you to have a flow of time, a sequence, an agreed upon shared state for any specific point in time, and for certain things this is necessary, but you can use other data structures too like tangles(DAGs) or even what SAFE Network is doing, depending on what you need to do.
Steemit needs the data structure it has but you can do sidechains with it to extend it.
The benefit of blockchain is that you can bake in economic incentives to reward wanted behaviour. Or am I wrong?
Wow, actually... That's brilliant I hadn't considered the economic incentive before.
@markopaasila this gives me a brilliant idea for something completely unique to steem.