Distributed Trust Systems and the Kenyesian Beauty Contest
By Joel Dietz - email@example.com
How distributed trust systems can solve the Kenysian beauty problem and create a more efficient startup ecosystem
One of the most persistent in modern market systems is the Kenysian beauty contest. In short, when asked to pick the most beautiful person, the most profitable and efficient strategy is to predict what other people will predict, not attempt to converge on what is actually the most beautiful. Modern market systems inevitably follow this trajectory.
A distributed trust ledger may follow a different logic than a market and create other tracks complementary to it. Numerous "social web" systems in which karma or other points are accumulated show divergent logics. Example services are Amazon reviews, Wikipedia articles, Facebook likes, Facebook friends, Twitter followers, Instagram likes, and aggregation services like Klout. In many cases, there is a very significant divergence between people who have some sort of accumulated social capital via their contributions and financial capital.
Consequently, accumulated and now quantifiable social capital has begun to take on financial value. Systems like Klout offer perks to users with high enough scores. This can effectively "unlock" certain types of services. To a certain extent, this is similar to what the credit scoring agencies already do for the American credit system in a rudimentary way that does not take full advantage of this larger societal shift towards a more connected world.
When it comes to market systems, these basic principles can be applied. J.P. Morgan famously quoted, "a man I do not trust could not get money from me on all the bonds in Christendom." Earlier trust systems which to a certain extent governed markets and abuses thereof were effectively managed by internal nontransparent trust networks within the industry. This self-regulatory principle often allows greater efficiency and trust than systems which require an external regulator (which frequently have a single point of failure and can be gamed). Other fascinating historical examples were the monastic networks which were instrumental in the founding of the banking system.
However, one important innovation on previous trust systems is that not only that do not necessarily require a single arbitrator. If as with Bitcoin or other blockchain systems, the fundamentals are transparent, then the actual interpretation of that data is open to anyone. Consequently, decisions on how to interpret the data are driven by a single use case, rather than a single interpreted number (as in the case of current systems, such as Klout).
A trust system which would allow more efficient and effective markets would likely need to have the following characteristics:
(3) some degree of anonymity or pseudo-anonymity
(4) complementary systems which 'gatekeep' depending on presence in the trust system
These techniques already exists in many sectors. It often makes sense to gather an amount of social capital and notoriety by providing some public good, and then capitalize on it by creating a private good which is profitable. The Sean Parker story from Napster to Facebook is an excellent example of this trajectory. One advantage of this, which is found in traditional startup literature, especially that of YCombinator's Stanford class (CS 183), is that this often leads to high value cohesion among the initial team. In short, the joint desire to provide a service without a significant financial incentive leads to cohesion in the team then allows them to continue to cohesively execute even when there is a significant financial incentive.
This is why some networks even deliberately weed out people who are hanging around for the potential financial incentive and instead encourage volunteer efforts or ask extreme vetting questions like "would you do this even if you had a year left to live" as in the AirBnB case. However, like all volunteer based networks, this has the downside risk of decreasing the amount of effort over time as finances remain a baseline need for human existence.
This multiple track technique toward assessing economics, appropriately called Swarm Economics, is not a new thing. It derives among other things from the efforts of the Santa Fe Institute to create modeling that was scientific and based on our increasing knowledge of natural systems (for example, the various types of stigmergy that appear in ant colonies).
As is now well documented, trust has biological and chemical correlates in the human body that can show just how well trust one is to their current environment. There is also a growing literature around neuroeconmics that shows that high levels of trust are highly correlated to high levels of productivity.
All of these movements, especially emergence of robust distributed ledger technology (Bitcoin, Ethereum, IPFS), our increased knowledge of human chemistry, and our ability to model economics at multiple layers and assess the relative efficiencies of these systems, suggest that we are in a moment when distributed trust systems will allow us to solve the persistant Kenysian beauty problem and create efficient ecosystems for promoting innovation.
Alexander, V., Tripp, S., & Zak, P.J. (2015). Preliminary Evidence for the Neurophysiologic Effects of Online Coupons: Changes in Oxytocin, Stress, and Mood. Psychology & Marketing. 32(9): 977–986.
Crosby, L. A. & Zak, P.J. (2015). The Neuroscience of Brand Trust. Marketing Management. May: 22-23. [Digital Version]