A Schelling (or Focal) Point is a solution people intend to use in the absence of communication, because it seems natural, special or relevant.
Schelling Points Are Everywhere
It’s a fascinating concept. Before mobile phones and the web, Schelling Points played a key role on how we self-organised locally. Churches, community centers, and bars would all serve as a Schelling Point. In fact, one could argue that culture in itself is a grand game of continuously shifting focal points, where people tend to conform towards the expectations of others. For example, imagine two prisoners held in separate rooms. Each are given the same sequence of numbers, and if they choose the same number, then both are set free. The numbers are: 183630479326, 73672, 100000000, 4782721. Which one will they choose? In order to win, the participants need to assume what the other is thinking. And thus, both would likely choose 100000000.
It’s A Coordination Game
But now comes a more interesting meta question: how can you make parties converge on a mutually consistent decision framework WHILE in the absence of communication? There are many tools in our arsenal that help us solve coordination games.
For a while, I’ve had this vision of creating some design that will allow all networks to mint their own value. It will be organic & automated. We will unlock HUGE amounts of social & other implicit capital: adding a lot of new economic liquidity & agency.
This thought initially branched off after seeing Dogecoin, a meme, become a currency . After realising that its value was,in some way, correlated to the network & attention it garnered, I saw this as a way to define an “in group” to the meme itself. The next step was to ponder on if we could all just mint personal IPOs? Where its’ value will be related to the network effect we have on the world. I envisioned a world where everyone around me were allowed the opportunity to have access to new economic liquidity through such a scheme.
However, my gut feeling told me that none of this feels elegant (yet). Creating a new altcoin for a new network effect doesn’t systemically quite add up.
I tried this when I played around with The Cypherfunks: a decentralized band, where we created an altcoin in early Feb 2014. The altcoin was simply supposed to represent the network (the Schelling Point) and attempted economic gravity well. Additionally, if one were to just IPO various networks by oneself (i.e. a person or a new community), it’s not only likely illegal, but also systemically ineffective (not even considering thoughts about death). In a global world,when you want to coordinate around common goals, ideally, you don’t want its existence to be reliant on a single entity/person to will it into existence. There could be several people around the world thinking: “We should solve problem x”. But in order to get a network/organisation off the ground to move towards solving “x”, would require using the common tools of organisational structures to do so. There’s an immediate barrier to entry.
The Protocols Guide The Way
So the requirements became:
Be able to allow any networks & attention attractors to start accruing value if there’s interest in it, WITHOUT a central party. This allows people to start coordinating by only knowing where to look.
Programs themselves can coordinate without appealing to human insight or knowledge (out-of-bound).
Anything that gathers interest and attention can have a network of value: From people to things to ideas. From Simon de la Rouviere to Apple to Doge.
It felt like there were clues everywhere. I just had to keep digging.
Organisations itself felt like a subset of a type of system that creates these incentives to coordinate (in a pre-internet way). By having a common goal as the Schelling Point, people can work more easily together vs. the market itself, because there are various implicit behaviours that does not have to be dealt with. Things like, having implied trust (i.e. your colleague won’t kill you) & heuristics around decision making that are similar (“i.e. we optimise for x”).
Maciej Olpinksi started writing extensively about the concept of attention markets, and that influenced my thinking. Perhaps in a superset: organisations are attention attractors themselves (memes all the way down). The meme of the organisation is glue. I experimented with ideas like: Attention Bonds, where an actor earns interest in the form of attention in a person. However, it didn’t fully yet feel like the best design.
I also started digging back into my masters research, where I studied information overload and how individuals & groups deal with critical mass. As long as there is novelty being produced by a group, it attracts new members. As it grows, marginal costs starts to reduce and depending on the design, it reaches a point where no more novelty can be reasonably produced. At this point, user churn occurs due to noise.
Last August 2015, I wrote a paper for the Hong Kong COALA Blockchain Workshop titled “Maximizing Novelty: The Potential Use of Blockchains in the Design of Sustainable Online Communities”. It centered around the idea of how blockchains can help solve the issues of information overload and critical mass. It rested on the premise that to keep novelty high, interacting with an interest network will become automatically more costly over time. In this way, an IRC channel, still retain relative novelty at critical mass.
I didn’t realise until last week in Shanghai, while speaking with Maciej again over whiskey, that the ideas were most definitely correlated. In fact, I feel a bit daft that I didn’t make the connection earlier (even though the pieces were already there).
Could there be some design that represents the value and interest of all information? After digesting the various components, I feel I’ve come closer. I want to introduce a concept that allows:
Easier, obvious Schelling Points for coordination (globally namespaced).
Ability for all information (“topics”) to automatically have its own network of value.
It being produced without a central party (only governed by a protocol).
Not disproportionately reward early-adopters (IPOs reward those who are at the right place at the right time with the right amount of money).
Allow reasonable investment/speculation into topics (betting that said topic will get more attention in the future).
Allow each network to accrue in value when more people care about it.
Add skin-in-the-game value signals to information.
It’s a simplified variant of the Persona Attention Bonds with a different issuance scheme. It’s designed and built to be run on Ethereum, a global decentralized blockhain world computer.
The Protocol Defines The Cost
On a global namespace, in the form of a #hashtag each topic has an attached, algorithmic coupon issuance rate.
To issue a coupon, one pays with another coupon such as ETH, for example, to mint a coupon based on the rate the protocol determines. This coupon is then dispensed for actions related to that topic. The cost of the coupon changes depending on interest in the coupon. These cues come from information overload research about how this cost of the coupon should change. In a certain period of time, it’s cost will grow exponentially, upon which, if new coupons aren’t bought, the cost of the coupon will linearly decay. More thinking needs to be done if these are the optimal curves. The coupon can be kept, transferred (and thus bought/sold) or immediately used. The decay makes sure that early adopters don’t just buy up all the future attention, and thus allows the topic to attract interest dynamically at lower costs when others would want to care about it.
What the coupon can be dispensed for is up to app developers to make & design. A simple app is to dispense the coupon to make sure that those who are interested in that network, can signal what others should care about it. For example, posting a link or text costs 5 coupons and upvoting a link or text costs 1 coupon.
You might wonder to whom do the funds go to when buying these coupons? For each coupon bought, the buyer can choose where the funds go. It’s an additional signalling and coordination tool. The default would be for most topics to simply burn ETH. Burning implies a proof of sacrifice to be part of that “in group”. However, in some networks, having that ETH sent to someone or a group makes that network more valuable. It also means multiple groups can remain coordinated on the same topic before it is necessary to potentially split off into a separate topic.
A Personal Attention-Action Network
Each human has limited attention, time, and resources to spend. Others can vie for a human’s attention. In an ideal setting, coupons are bought and paid for to the person. Coupons are expensed to add tasks & upvote tasks. The outside world signals to the person what they should be spending their time on. If the person complies, more people will buy coupons for their gain.
For example, imagine that Vitalik Buterin (#vitalikbuterin), inventor of Ethereum, with the following tasks:
231 upvotes — Research Casper & implement PoS.
201 upvotes — Check this funny doge meme and tweet about it!
131 upvotes — Donate your money to longevity research.
If you think someone’s attention will be valuable in the future, you can buy coupons now to use later, or sell later.
If the person dies, there can still remain a residual attention market around the person’s life & ideas. Thus, a lot better than a personal IPO and expecting to get returns on future cashflows.
Modelling salaries could also work in this regard, and adjust dynamically as a worker’s skillset improves. For example, ConsenSys being my primary employer would spend a certain amount each month buying up my coupons to upvote a task that says: “Work on ConsenSys projects.” Other players (i.e. co-workers) can then solicit my attention as an equal player to ConsenSys.
Imagine a world where where programs and advertisers start buying and paying you for your attention.
An Open Source Organisation
Imagine Truffle (#truffleframework), a framework that makes Ethereum development easier, buying coupons that allows developers and community members to signal & vote on issues related to the tool & its’ future direction. The funds are sent to a communal pool where either the whole community can solicit developers to come work on Truffle, or a group of directors can decide.
The more useful the tool gets, the more its coupons become valuable for actions. You can have additional possibilities here, for example, submitting a pull request costs a coupon.
Global Warming
One of our biggest collective action problems is: global warming & climate change. Hashtag markets hopefully can produce incentives in the following way:
A Schelling Point for coordination #globalwarming.
Investors can produce attention directly.
Those who do not know how to help can simply help by buying attention. If the value of the coupons keep rising, it will attract attention in and of itself. For instance, an investor spending $10 m consistently buying up #globalwarming coupons will hopefully keep the attention flowing. See it as a donation towards the idea, and to a specific person or organization instead.
My Thoughts on The Future
As you can see, one can create a market around any meme. Some memes afford different opportunities (like the OS organisation part of voting on issues).
The biggest assumptions at the moment are:
The value of having easier coordination vs. the cost of doing so. Free coordination in the absence of a Schelling Point might be good enough that people won’t bother. The cost is more than the benefit.
The 2nd biggest hurdle is designing the optimal graphs for coupon issuance. My gut feel is to take cues from current existing resource models and actually from (strangely enough) whirlpools. When it is accelerating hard, start increasing costs exponentially, and then if acceleration decreases, implement a linear decay.
This model still relies on people having an existing token (say ETH) to exchange it for the coupons. This still creates a barrier to entry to engage. Even IF everyone in the world has easy access to ETH, it is a type of barrier. Ideally, you would want people to participate and gain from being a part of these networks with only directly applying their willingness to part with their attention. There are cues here from projects like Benefactory: essentially allowing a community to pledge together to fulfil debts. In other words, given historical proofs, one could exchange something else (one’s reputation to fulfil pledges), for coupons.
The other question is the value of one’s bag of coupons and contribution history. Should cost decrease to buy coupons if there’s consistently useful, novel behaviour/value in that hashtag?
When We Lose Abundance, We Start Distrusting Each Other
My gut feel is that this format could allow for a lot of currently unknown emergent behaviour. Much of this thinking does not only come from my interest of wanting more agency for individuals & the ability to coordinate on scales we haven’t seen before. But it’s also driven by the realisation that our current economic system is wholly unfit for where we are and where we are heading. Automation will continue to come & none of our current economic models will help us retain social liquidity.
I’m pretty sure this is a classic zeitgeist mind meld happening at the moment. Much credit especially goes to Maciej Olpinksi & Meher Roy, who continue to think in this direction. We are currently sharing concepts and models, which hopefully coalesces to something useful in the near future! :)
As Meher recently said, “The intrinsic value of a shared interest network token equals the net present dollar value of its associated (monetizable) attention flux.” Spot on!
Share your thoughts!
Comments