The Economics of Big Data
As we move closer and closer towards the vision of the now cliched ‘global village’, trendier sections of the global citizenry might reckon we need a more ‘happening’ term to describe our increasing interconnectedness. Revolution 4.0 presents a landmark step in increasing the degree of this connection, and instrumental to this is big data.
An integrated, connected society will have some fascinating outcomes for enterprises that manage uncertainty. As our predictive capabilities improve, the economics of amortization worsens.
Big data is essentially in the business of predictions. It aims to minimize uncertainty by identifying patterns and trends, fitting to a curve, and through the science of correlation. If we can leverage the power of data to foretell the odds of you having an accident, contracting the COVID virus, or the likelihood of you being in a plane crash, we can take steps to reduce it. Big data provides us with the cushion that humans cannot live without – humans don’t like uncertainty. They don’t thrive in ambiguity. Instead, they interpret the physical world and the information contained in it to understand events, and increase favorable consequences to them, as a race.
Reflect on this thought for just a second. Segments of industry are devoted to amortizing risk across populations. Insurance is a good example, socialized medicine, yet another. There are several others that are noteworthy: travel ticket rates, credit, agricultural trends, and so on, and so on.
A mortgage is rather morbidly, connected to death. Let’s trace the etymology of the word, and you’ll understand that the AICoreSpot team isn’t trying to be ‘dark’ for the sake of it. The word traces its roots to the French word “mort” and “gage”. A gage is something of an undertaking. And mort refers to the literal expiry of the undertaking, it means death; it will cease to exist when a specific condition detailed in the mortgage is fulfilled.
Prediction and amortization are two polar opposites. If you’re blissfully unaware about a population’s risk, you utilize a socialized medicine model where all are subjected to tax equally. Insurance operates on the same principle. Risk is spread out. If you have more knowledge, you have variations in pricing based on current risk factors such as drinking, and smoking.
In any such model, even with the so-called sciences of prediction, and correlation, there is an aspect of ambiguity. Within the subgroup of, for example, stunt riders, or drug users, you really have no way of knowing who will die in a fatal accident or of a terminal illness.
Therefore, the ideal prediction? An insurance policy custom made for one individual. If we knew for sure the odds of us having an accident and the financial consequences of it, your monthly insurance payment would just be periodical deposits into a bank account which would eventually equal the monetary consequences of that accident on the day of the event. And, obviously, the insuring individual’s administration and profit margins.
A tax on your mortality. Monday morbidity.
Of course, no predictive framework is bulletproof. However, as we leverage data to make increasingly robust decisions on the basis of updated data, (owing to the IoT world, and Bayesian probability calculations) things start to get a tad bit weird. In that terrifying millisecond prior to a collision, as the driver jams down on the brakes in desperation, we can estimate the probability of their being in an accident with almost 100% accuracy. Do we terminate their coverage or increase their premiums?
Basically, businesses that handle reducing risk are faced with the near-certainty of a more connected, calculated, and predictable universe.
Economists have long engaged in discourse about the “ideal market” of a commodity good in which supply and demand rates mandate pricing for long time frames, but we are well aware that this is not a very practical model in real world scenarios – it doesn’t offer much in terms of utility. Branding, human fancy, and bias have a far bigger influence on price elasticity and market share than we could every fathom.
At a certain stage, arbitrage markets will face deflation due to data. The uncountable money with regards to taxi cab medallion speculation in NYC – in some scenarios, the fortune of a family being passed down from part to child – are being quickly reduced in value by a business like Uber which eradicates the taxes a taxi dispatcher could collect.
We might soon wind up several economic theories that have been taken for granted as being true when we are able to collapse the inherent risk in a business leveraging data. Efforts such as lean startup are basically a bid to collapse the risk up front, resulting in reduced reward for future investors as the certainty of market demand is present.
Let’s take a look at one more scenario of risk eradication: Kickstarter. The organization has already pumped 200M US$ to new projects, and is expected to up it by another 500M US$ by next year. However, not one of these projects obtain funding until they have established two things: one is customer demand, and the other, an enticing message. Therefore, the time-tested build-it and see if they bite model – and the ROI and profit margins required for justification of the consequent risks – is to a degree, dated.
What are the implications? As we increasingly improve at the science of predicting consequences, we’re have a lesser likelihood of pooling our assets owing to inherent uncertainties? And if so, is a world that we can predict, an individualistic, libertarian dystopia?