I remember a study of mathematician Odlyzko, that the value of the internet comes from the interaction and not from the content anyways. Steem is nowhere near rewarding the most communicative pieces of content.

In this is why we don't pay attention to studies by mathematicians when it comes to understanding value. Let's take, as an extreme example, the idea of a network with no content on it. What is the value? There is none – unless the content is just the list of transactions. But then, there is content! And only the content has value!

Value emerges from content intersecting with individuals. The process of that intersection is interaction, but to suggest to that only the interaction gives rise to value is the take away the wrong idea. Content in and of itself has no value. Users, in and of themselves, have no value. The process by which they discover content, experience content, and distribute content gives rise to value. Absent content, there is no value to give rise to. Only with content is there a context for interaction to occur.

Now, if you take away from his study that you have to construct the entire chain from the creation of content to the discovery of content by an agent to the acquisition/experience of that content by an agent in order to have value – then you've come away with something useful. That provides actionable knowledge. You can concentrate at any point along that chain and provide tools which improve the flow and thus increase value.

I know such localism only from systems like Scuttlebutt and Urbit but more from an infrastructural point of view. I can barely imagine how value distribution happens in such a patchwork environment. As far as I know systems like Utopian with bottom-up established networks are not really used by people - "too cypherpunk" for most I guess. And it lags the big, attractive honeypot which blockchains are.

Value distribution occurs by discovery and reward, just as it does through every other market. Including those markets which are not free; even in unfree markets, value distribution occurs by discovery and reward, it's just not very efficient and often not very effective, and ultimately ends up wasting more value than it discovers and thus diminishing the entire system. The patchwork nature of environments like Scuttlebutt give rise to different means of discovery. Someone whom you explicitly designate as a trustworthy source, i.e. someone you follow, or someone they consider a trustworthy source, i.e. someone they follow, writes about/interacts with something they find valuable and you hear about it. Then you do so and people who may have been outside of the informational horizon of the first person now have the ability for that discovery and thus that value to happen. In theory, SSB is also supposed to gain discoverable value by hyper localization. In its standard mode, you should be able to see other network participants who are in your literal physical locality, engage with content they produce or themselves engage with, and thus open up that gateway to the process of valuation discovery beyond the limited horizon.

In a very specific way, hyper-local discovery networks transmit value in exactly the same way as Facebook or Twitter. They act as a communications substrate over which human interactions with content can be broadcast and observers can decide if that interaction or that content is sufficient to motivate them to act.

"Blockchains" aren't really big, attractive honeypots. They are hype centers. Very few of them have actually been used to solve a real problem which people have. They are slow, bulky, awkwardly deployed databases with some inbuilt redundancy like a late 70s era VMS cluster. And that's it. I've worked on both.

Without reward-pool, you have pocket-to-pocket tipping, I guess. Otherwise usage of such a system would be one big "disagreement on rewards".

Without a reward pool, you have no inherent inflationary pressure, no inherent push to treat other creators in the space as if you were sharing a "crab pot" (where everyone independently observes that in order to maximize their opportunity it's in their best interest to pull down anyone who starts to crawl to the top, and if you find yourself at the top, it's in your best interest to make sure that no one else joins you there). Tipping acts as an important signaling mechanism between the user and the system because the user actually has to sacrifice something in order to make that signal happen. It doesn't have to be a significant sacrifice but the psychological value is disproportionately strong.

There can never be "agreement on rewards" in a real sense, no matter what your system is, because every element of content is not equally valuable to every consumer. This is an essential failure point of systems which attempt to surface content based on what "everybody thinks is good." It doesn't matter what the underlying mechanism of the expression is, the essential assumption is wrong and so the ultimate result can never not be wrong.

I know similar discussion about a individual "decide what you want to see route" vs. authoritarian "this is what you are allowed to see" but haven't seen any concrete decentralized solutions.

We used to have a fair amount of experimentation with web of trust systems, and a lot of that research ended up getting rolled into advertising architectures rather than user facing experiences (though under the hood, the Facebook timeline filtering algorithm and of the Twitter "best timeline" algorithms oh a massive debt to those original individual-focused preference-signaling methods). Advertisers are absolutely interested in putting something that they have that you might want in front of you because you might buy it. Contrary to popular belief, they don't actually want to waste their time or yours in trying to put something you don't want in front of you because they know that will just annoy you and waste their money.

Social media platforms ended up developing a top-down, authoritarian philosophy because a lot of the developers who ended up doing most of the work actually have that as a personal philosophy. "Everybody would be so much better if they all agreed with me because my opinions are clearly right and better, and since everyone I know agrees with me, that's obviously the best way to determine what ideas are good."

There are some good reasons I don't get along with my peers and that basic social assumption is one of them.

It's also very, very wrong.

I would love to see more platforms take an individual focus with spreading web of trust quanta as a lensing mechanism. I've written about at least one way you can implement something like that using what is now fairly traditional blockchain tech, but there are other, better, distributed methods. The question is if any of the platforms really want to let go of that amount of power and decentralize the locus of control to that degree.

To be fair, the Scuttlebutt people are no better about that than anyone on Steem or Hive. Back when we started see a lot of pushback on the more popular social media platforms and Gab moved onto a Mastodon federated instance (and one day I will go on it disturbing length about how the people on Mastodon went from being free-market socialists to hard-core authoritarian free-speech deniers in about 24 hours), there were high-level discussions among the Scuttlebutt literati about how they could lock down their extremely distributed, hyper-local, designed-to-be-proof-against-censorship mechanisms against wrong think and against "bad people" with nary a microsecond of self-awareness.

The most promising folks I'm aware of are working on LBRY and recognize that to have a true, useful discovery mechanism that they will have to implement a web of trust system sooner or later. Right now valuation/search ranking is based purely on how much in the way of tokens are "staked" on a particular channel or piece of content, but they recognize that solution is not the end-all be-all and has some significant problems, and they want to move to a more individual-focused signaling system going forward. That's about the best news I have on that front.

It's kind of a sad world.