Donella Meadows’ book “Thinking in Systems” is a masterpiece that offers a different way of thinking about everyday problems. We are moving from a data-scarce world into a data-driven artificial intelligence era.
In her book, Meadows had highlighted issues in relation to data reflecting the reality. The “Fifth Discipline”, a book by Peter Senge, is another good effort that sheds light upon similar topic. Between the two titles, I however would give more marks to Thinking in Systems.
What is a system? It’s a set of inter-connected bodies, which interact with each other in a specific way. The system has capability to make self-adjustment, self-organization and evolves by its own. Generally speaking, systems are sustainable, and they would recover automatically after temporary disruptions.
When a system encounters external shock, it may react in different ways and at different speed. A complete system is made up of three key points, elements, connection and target. For instance, Alibaba’s online shopping ecosystem consists of sellers, buyers, courier service providers, platform operators.
All of them connect with each other in order to match demand and supply sides, moving goods from seller to buyer to complete a transaction.
Online vendors with good reputation can easily expand their sales volume. A positive feedback loop will place the sellers at better position on the website. Other less popular sellers have to ramp up their sales and try to get more good feedback.
Because sellers with good rating are extremely popular, we have seen a new profession being created — click farmers.
They are hired by online sellers to artificially boost “likes” and positive views using hundreds of cellphones and several hundred thousand SIM cards.
Such practices increasingly tarnish the rating system of the online shopping platforms. To safeguard the credibility, platform operators react by cracking down on sellers involved in click farming.
The rating system then restores its credibility, until click farmers figure out new tricks to again generate fake likes.
Similar murky practices must have taken place in the past, but in the Internet era, online data can be collected more easily to tell the truth and uncover any wrongdoings more speedily.
This article appeared in the Hong Kong Economic Journal on Jan 18
Translation by Julie Zhu with additional reporting
[Chinese version 中文版]
– Contact us at [email protected]