The title of this blog is a play on the famous words by the pioneering computer scientist Alan Kay who said that: “The best way to predict the future is to invent it”. In contrast, the ‘Standards for standards’ approach seems to take the view that: The best way to predict the future is to standardize it. In other words, it seems to predict future directions for innovation and preemptively create a standards template for future innovation.

This approach will not work for a number of reasons, but in this post I will focus on two specific issues:

Firstly, That future standards scenarios are truly unpredictable since they are cross domain. And secondly, The ‘standards for standards’ approach goes against the philosophy of the Web i.e. Long Tail.

Background

Let us consider some background:

There are two ways to create standards: The traditional approach which worked in a simpler, product driven environment, was based on the standardization body (often managed by national governments), deciding the standards in specific domains. This proved to be very cumbersome and inflexible especially as the rate of change increased and western economies moved from being product based to service based. The rise of Information technology also led to an increased rate of change. To overcome these problems, we saw the emergence of industry consortia. By definition, this led to a reduced emphasis on the older ‘State controlled’ centralized standards setting approach and leant more towards industry working with other players to solve practical problems. The consortia led approach (which dominates currently), depends on deploying specific innovations which had a proven business model. Since the potential to innovate is infinite, then by extension, so should the ability to form consortia around new innovations. That’s why the current consortia led approach works well.

Now, the ‘standards for standards’ approach attempts to go back the older, centralized model by creating some kind of ‘class system’ between consortia – with some standards being more equal than others based on an arbitrary set of criteria. There are some obvious drawbacks to this approach such as: How do we decide Quality? Which business requirements/use cases drive the ‘standards for standards’? How do we balance between good enough vs. perfect? How do we make the ranking subjective and fair?

Apart from this, we will hit far more complexity in future.

Not only will ‘standards for standards’ be hard to do in the existing scenario, it will be MUCH harder to do going forward because much of the new innovation is likely to be cross-domain, i.e. it spans previously traditional domains. For example: Smart grids, e-health, Mobile health, Internet of things etc. All of these span existing domains. In the case of Smart Grids, the domains are Energy, IT and Telecoms.

What ‘standards for standards’ apply to the Internet of Things?

Here is an example: How exactly will we go about standardizing the ‘Internet of Things’?

The paper Device interoperability: Emergence of the smart environment ecosystems (pdf found here) discusses the question of interoperability for a smart environment and provides some interesting insights. Smart environments and smart spaces are a widely studied concept in ubiquitous computing research, ambient intelligence research, and Future Internet research. In spite of extensive research over the last 20 years and extensive government funding, the idea of Smart spaces/Internet of Things has remained localized to specific scenarios (ex monitoring tides). The problem is: Everyone agrees that there are benefits, but no one can point to a simple business model. The basic idea has even been proved in Internet, where the cloud computing paradigm, has been extremely successful. But in physical spaces (Internet of Things), no real large scale smart environment ecosystems exist despite vast amounts of investment and research. The paper takes a different approach and believes that the reason why these ecosystems are missing is that current technical solutions do not fit into the business interests of the industry.

As shown in the figure below, device interoperability can be divided into three separate levels: The physical level, the service layer and the information layer. In a smart environment all the three levels must exist and the ecosystem is extremely heterogeneous. The variety of solutions is enormous and it is impossible to support them all or even a small subset of them in an economically feasible way.

So, where do we start in creating a ‘standard’ for Internet of Things standards? The above scenario is an example but will be more common going forward in many domains. Are we seriously expecting governments to anticipate and standardize all these situations no matter how well funded?

Not Long Tail

Secondly, the ‘standards for standards’ argument goes against the Long Tail philosophy which drives the Internet. In a nutshell, “The Long Tail or long tail refers to the statistical property that a larger share of population rests within the tail of a probability distribution than observed under a ‘normal’ or Gaussian distribution” (from Wikipedia).

So, imagine if Jeff Bezos of Amazon sat down to set a ‘standard’ for ‘rating books’. Do we see that happening? How do you compare the ‘Lord of the rings’ with ‘Harry Potter’? The answer is – you don’t!

You let the market decide which is exactly what Amazon is doing through Amazon reviews.

To conclude:

I have an amateur interest in Quantum mechanics cultivated through reading books from Stephen Hawking and others like ‘In search of Schrodinger’s cat’ . Hence, a quote from Niels Bohr (1885 – 1962) often regarded as ‘as one of the most influential quantum physicists of the 20th century’ seems an apt conclusion.

Niels Bohr said: “Prediction is very difficult, especially about the future”.

If a genius like Niels Bohr could not attempt to predict the future, I think nor can we…

And if we cannot predict the future, certainly we cannot attempt to standardize it preemptively!

Image source: Device interoperability: Emergence of the smart environment ecosystems Version 1.0 Juha-Pekka Soininen et al