PsyFi Search

Monday 24 April 2017

Putting Pro-Innovation Bias on the Blockchain

Blockchain Bias

Over the past couple of years I've spent a lot of time listening to people wittering on about "the blockchain". They've claimed it can solve a plethora of society's ills - everything from the elimination of poverty to the overthrow of fiat currencies and the nation state.

Being charitable this is evidence of pro-innovation bias in all its perverse glory. Being cynical it's evidence of people trying to scam investment funds by capitalizing on the halo effect. The blockchain is a brilliant piece of innovation, which will one day - probably - lead to significant economic benefits, but in the end it's just another piece of technology.

Ledgers, Everywhere

Blockchain is most commonly associated with Bitcoin but is, in fact, a separate innovation upon which the world's least efficient payment system is built. Technically, blockchain itself is a single example of a broader class of systems known as Distributed Ledger Technology or DLT. 

A ledger is technology that's been around for a thousand years (see: In the Beginning Were the Accountants) - it records transactions so that you know who owns the various assets it records (including money). A distributed ledger simply means that the same ledger is copied to all of the people who use it - so imagine every bank in the world using a copy of the same database. 

The key idea here is that the entities using a network underpinnined by DLT don't need to trust each other, but can still trust the information on the ledger. This is, obviously, counter-intuitive, but it's a revolutionary idea.

Byzantine

The genius of the technology is in something called a consensus protocol, which is the mechanism that allows the users of the ledger agree to put a new record on it without trusting each other. In Bitcoin consensus is achieved through "data mining", which is a solution to a problem in Game Theory (see: Games People Play) known as The Byzantine General's Problem, originally described by Leslie Lamport, Robert Shostak and Marshall Pease.

In the problem a bunch of generals are surrounding a city and must decide whether to attack or retreat. Their challenge is how to come to a consensus on what to do when they only have remote communication and some of them may defect and start lying. Data mining via proof-of-work is the Bitcoin solution, but that's just one example of a DLT consensus protocol.

Owning Assets in the Interweb

What DLT does is allow us to know who owns any digital asset. For example a Bitcoin is just some ones and noughts. So you could copy it, use it to buy something from Amazon and then use the same Bitcoin to buy something from eBay. But the ledger records who really owns it, so eBay checks the ledger before they send you your pink tutu and finds that you've already transferred it to Amazon in return for a sparkly tiara. So no tutu for you you. Boo hoo.

However, there are a couple of things that DLT is self-evidently not much use for. Firstly there's rarely any point using a distributed ledger across a single organisation. They're expensive to run and synchronize and also pointless if you can just have a single bloody great database that you can control access to. Secondly they don't allow you to really track physical assets. You can put a representation of the contents of a shipping container on a ledger, for sure, but you still only know what's really in the container by having someone trusted look inside it.

Hype, Over, Case of

Despite the limitations this hasn't stopped the floating of mad ideas for use of the technology. One of my favourites was this: "Blockchain technology trialled to tackle slavery in the fishing industry".  I mean no, just no.

The willingness of entrepreneurs, technologists and hopeful fund raisers everywhere to push DLT onto every imaginable use case is a case of pro-innovation bias. The concept arose out of research into the diffusion of innovation by Everett Rodgers, who describes it thus in Diffusion of Innovations:
"The pro-innovation bias is the implication of most diffusion research that an innovation should be diffused and adopted by all members of a social system, that it should be diffused more rapidly, and that the innovation should be neither re-invented nor rejected."
Basically it's the tendency of people to over-hype the use of a particular technology while ignoring its weaknesses: innovation is good, we should adopt it. Blockchain is simply the latest example, but anyone who can remember the late 1990's will recall that the Internet was similarly hyped. Before that the great hope was nuclear power and at the beginning of the 20th century it was radiation.

As you can imagine, not all of these innovations turned out to be unalloyed successes when deployed unilaterally.

Diffusion Theory

Rodgers' diffusion theory is a way of explaining how and why innovative ideas spread through society. It's where the memetic ideas of early adopters, laggards and the others originates from and it draws on ideas from multiple disciplines - in particular it's a social mediated system, dependent on communication channels and opinion formers. This is similar to the concept of self-enhanced transmission bias that we looked at in Facebook Friends Can Make You Poor.

Tom Nicholas, in Does Innovation Cause Market Crashes? has suggested that bursts of pro-innovation bias can trigger the stockmarket surges that predate market crashes. His analysis indicates that innovation was at least a partial driver of the market run-up prior to the crash of 1929. 

Saints and Sinners

There is another possible explanation for the proliferation of stupid DLT use cases: enterprising individuals looking to raise investment funds on the back of the halo effect (see: The Halo Effect: What's In a Company Name?). This is the same behavioral bias that advertisers capitalize on when they use celebrity endorsements, recognizing that a popular celebrity associating themselves with a brand can cause the fans of said celebrity to rush out and purchase the product. Although the chances of a famous model actually using a consumer brand cosmetic must be quite low, I'd have thought ...

In truth these biases likely have a positive impact on overall human technological development. All nascent technologies take far more time and money to develop than anyone ever appreciates - to the extent that if we did realise this to start we'd never actually start doing the research. All of the money wasted on dotcom technology and fibre cable in the late 1990s eventually turned into the online world today. But the winners and losers weren't evident at the time, and only became clear much, much later.

Innovation is not straightforward, and it's often not clear until afterwards whether any particular innovation will turn out to be useful or not, let alone whether it will revolutionize life. The idea behind the blockchain, that we can establish trust in transactions carried out in a network populated by entities that don't trust each other, is more than clever technology, it's a new way of thinking about the problem.

And, in the end, that may well be the most important innovation of all.


Pro-innovation bias added to the Big List of Behavioral Biases

Related Posts
    

No comments:

Post a Comment