Why and when Big Data leads to a system crash

TLDR: In addition to great suggestions, the 2016 Communications Congress delivers an unpalatable prognosis: data-based control systems, whether in communications, technology or society, will inevitably get out of control. The whole problem and the prognosis, here in one picture.

The 2016 COMMUNICATION CONGRESS in Berlin was much more informative than DMEXCO in Cologne.

Reigned at the Digital Native naïve on the Rhine a deeply uncritical, even senseless "carry on like this", namely in the direction of "first artificial intelligence and then no more at all", there was no change among the communicators and Journalists influencers in the still structurally weak capital city a whole cornucopia of suggestions for digital thinking.

Problem: Big Data volume is growing 1.5 times faster than computing power

One of the highlights of a remarkable program was certainly the controversial keynote speech by Dirk Helbing from ETH Zurich, who warned of the socio-political consequences of Big Data.
His message: It is inevitable that Big Data will lead to a loss of control!

According to Moore's Law, the computing power of available computers doubles every 18 months. At the same time, however, the amount of data to be analyzed (big data) doubles every 12 months. The system is no longer controllable. It will most certainly get out of control.

Solution: None?

The system collapse will come. Anyone who hears something like that is worried.

Kommunikatoren und Marketer werden dann vor vergleichsweise kleine Probleme gestellt werden.
Ihre “Predictive Communication” oder ihr “Predictive Programmatic” wird ihnen um die Ohren fliegen. Das hört sich dramatisch an, ist aber unerheblich im Vergleich zu einem Big Data-Kollaps bei Infrastrukturbetreibern, z.B. in Verkehr, Energie oder Telekommuniktion (Google, Facebook <- bitte regulieren).

Of course, there are counter-strategies to keep complexity manageable. Data waiver, for example (the first rule in data protection).

But what if the fundamental computer law, Moore's Law, no longer applies?
When NO growth in processing power can be provided anymore and the data volumes grow anyway?

So let's hope for the quantum computer.