Skip to main content

Running (data) economy

29.01.2020
Gianfranco Cecconi
Opinion

The controversy around Nike’s “Vaporfly 4%” running shoes is everywhere in the news as I write this. It is claimed both by the makers and its users that the shoe design and materials offer an average 4% “running economy” compared with other top runners. Of course, if you’re a couch potato like me, the Vaporfly won’t make you a champion. But, if you’re a pro, such a feature can shave minutes off your marathon. Kenyan runner Eliud Kipchoge was wearing the Vaporfly as he achieved the standing official marathon world record with a time of 2:01:39 in Berlin on 16 September 2018. Now he flies at about 1:59:40.1 

Without taking anything off Eliud’s achievement, this has arguably become a problem. Athletics organisations are starting to consider regulating what characteristics make a running shoe acceptable to be allowed in competitions – the so called “shoe range”.2 It’s the same process that, for example, Formula 1 had to go through for years to ensure some degree of fair competition between racing cars in a sport that is already today accessible only to incredibly wealthy teams and the coalition of sponsors backing them. The outcome of this process may be that, someday soon, wearing technology equivalent to what is used in the Vaporfly would be considered like doping practices.

And because my mind is distorted by too many years spent studying and supporting Europe’s data economy J, as I listened to a podcast3 one morning, a parallel came to my mind: between the Vaporfly 4% and access to data as an enabler to technology development.

I do not think you need me to reiterate the instrumentality of data to much of today’s technological development. From simple application – where services are providing users with the right information at the right time - to the more complex and sometimes controversial usage of artificial intelligence, data is the key input to algorithms. And good data is difficult to find. Even considering just the “easier”, non-personal data published by governments for re-use by their citizens and businesses as open data, research4 shows how we are “plateauing”, even in the relatively advanced Europe. On one side, the EU Member States are consolidating the achievements so far, whichis good, but also slowed down by lack of enough investment and a general struggle to evolve the related skills and culture at a sufficient pace.

If we talk about data sharing in general – dealing with often more subtle, not open and definitely more delicate datasets that can include confidential information and personal data – it gets worse. It is estimated that in the USA the total cost of acquiring the data you need to build your AI – financial but also friction due to compliance or legislation in general – can be 100 times more expensive than in China.5 That’s a 99% “running economy”. Squeezed in the middle, Europe prides itself of values - and of the legislation developed accordingly, such as GDPR for personal data protection - that make acquiring data even more complex and expensive. That is also why initiatives such as the Support Centre for Data Sharing were started, in the attempt to remove as much of this friction as possible.

So, are we racing the same race? Nike’s Vaporfly are EUR 275 a pair. What’s the cost to Europe’s economy of the special “shoes” China is wearing? Should the World Trade Organisation regulate the “data range” different countries have access to for competition to be fair? Of course, it won’t. It is up to us to train harder, the old-fashioned way.

Running (data) economy
Image credit:
(C) 2005 Timothy Takemoto https://flic.kr/p/56qG1