I hate working out. But I dutifully drag myself to the gym a couple times a week. Recently, I’ve been going to Orangetheory Fitness. For those not in the know, “Orangetheory is a science-backed, technology-tracked, coach-inspired group workout designed to produce results from the inside out.” (OTF) You wear a heart monitor that is connected to the gym’s system and during your workout, you are served up data that shows how hard you are working. The goal is to stay in the orange zone which is optimal for positively impacting your metabolism and burning calories. After the workout, you are sent a report of your results. I’ll admit, I used my report today to justify having that coconut-macaroon cookie later on! I have wondered what happens to all this data and while there are privacy issues that could be explored, I want to take more of a philosophical, 500,000 foot approach.
Is the world just a bunch of data flows and are we just data processors? Am I, in some sense, just a compilation of all my data, both biologically, as in my Orangetheory data, but also in the data I construct by living my life?
Yuval Noah Harari’s last chapter of Homo Deus explores data-ism as the new religion. He sets out an argument that scientists have tacitly agreed that all phenomenon can be explained in terms of data and dataflows. In doing so, science has created a common language and master framework that allows disciplines as diverse as musicology and economics to share information in meaningful ways. From this foundation, he asserts that biology and computer science are both firmly entrenched proponents of data-ism, that organisms are just algorithms and both types of algorithms (organic and artificial) are data processors.
In Harari’s argument, we are made up of dataflows that are constantly performing data processing. However, compared to AI, we’re pretty inefficient at this task. We were built to process data 70,000 years ago on the African savannah not to deal with the vast amounts of data in the world today.
Harari makes a good case for the data processor argument and linking it to bigger issues like economic and political systems. He compares data-ism as a belief system which is replacing humanism, just as humanism replaced religion, as the dominant framework for explaining the world. However, it's evident he doesn’t want this to be true. He’s struggling with the tension of seeing all of the evidence pointing in a certain direction, but also holding onto the belief that there must be something else, that we can steer this ship in another direction. The last few pages read like a cry for help – for someone to point out what he might be missing.
So, if we are just “not so great data processors”, maybe the AI algorithms should take over? By which I mean, become the centre of power and control. But, can everything be reduced to data and explained through computations and algorithms?
I suppose we are testing that theory right now. Like Harari, I’m concerned for the human agenda. I wonder about our place in the system. Is our value situated in our ability to create and train algorithms? Once that work is done, are we no longer valued (or as valued)? Currently we are building systems in the service of humans but maybe there will be a day when humans are no longer required either to build those systems or provide the data, a day when the systems themselves will be in service of each other. On the other hand, if humans are not central to the equation, what would the smart fridge, iPhone, Roomba and my Orangetheory data want from each other? Yes, they can coexist as datasets to inform some type of algorithm(s), but to what end? It's hard to imagine from my limited human data-processing perspective.
Harari, Y.N. (2015). The Data Religion. In Homo Deus A Brief History of Tomorrow (pp 428-462). Toronto, ON. Penguin Random House Canada.