Getting Uncomfortable with Data
Daniel Tunkelang
3345

Data is the great democratiser

The greatest scourge of the twentieth century — and it is enduring into the 21st — was not sexism, nor racism but paternalism.

We lived in a two class world — a world where people appointed themselves as “the right people” to make decisions for others. We see this in the very word Government — govern is another word for control. We see it in the attitudes of EU officials — one of whom (Brendan Devlin) last week argued that the only way to get the energy system we want was to make it so hard for governments to get consensus decisions through that changes in the market couldn’t change the system over decades. And we see it in UK and US Universities who are educating millions of students to interfere in Africa and undermine all of their market systems in the name of “making a difference”.

Data is blowing a cold wind down their corridors of power. Suddenly people are becoming able to find out which Hospital Trust is actually doing a better job. Which BBC employee earns £235k to work from home on special projects (Janice Hadlow). Which government minister was hauled over the coals by the UK Statistics Authority (Harriet Harman) for misleading people.

These patriarchs are responding the only way they know how. By spreading fear, uncertainty and doubt about data. Shoring up their position by attacking everyone else’s.

Let’s take the self-driving car as an example. Anyone who doubts how much good this can do need only spend an hour in a supermarket car park. That’s right the bar on driving capability is set very, very low. And the sixties myth that driving was fun is hollow now too — only shored up by car makers and advertisers.

We kill 1700 people a year and have 160,000 reported accidents on the roads in the UK (one of the lowest figures per person mile in the world). Meanwhile self-driving cars have had 9 accidents — all while being driven by people. Yet we are awash with the self appointed czars telling us we must legislate against these killing machines.

We have the same in facial recognition. “Oh think of the harm” — put out by the same bodies who have installed CCTV cameras in every city, just to police bus lanes and parking (often, as in Reading, Berkshire, to force people into their off-street car parks and their cozy deal with a car park operator).

Slowly, gently, data is doing more for democracy than governments ever have. The Open Data Institute is opening up all the data governments used to go to great lengths to conceal. We are able to run our own algorithms and find out very different insights from those the public bodies wish to share.

We, as datascientists, are leading the way here. Datascience is moving from being a hard thing done by a very few people to a general public good, accessible by all. This is affecting every organisation — commercial or otherwise, as they are swept by a tide of transparent data to, at last, move to real decisions based on known facts rather than simply ego and the HIPPO (highest paid person’s opinion). The platforms for those who wish to control us and tell us what is “in our best interest” are crumbling too.

Our responsibility as datascientists is not to lose our nerve. It is us, not the little dictators and paternalists, who are building the better world for all.