Ascent Lookout 2014 Happiness@work – Spread a little happiness Zero email, how to get things done! Part three: The last mile, becoming Zero email Ascent magazine: A magazine into the future of our ever-more connected planet What is your future in this ultra-connected world?

The Importance of being Hugh Williams, or why you should care about patterns

On December 15th 1664 a shipwreck occurs near the cast of Wales; 82 people are killed, only one person named Hugh Williams survives. On December 5th 1785 a similar accident takes place in roughly the same spot; 70 casualties this time, and only one survivor, whose name was coincidentally Hugh Williams. Yet, on August 5th 1860, near Scotland, another shipwreck happens. All passengers died but one. Can you figure out the name? If you said Hugh Williams, you guessed it right.

Atos The Importance of being Hugh Williams, or why you should care about patternsThe moral of this unlikely chain of nautical disasters could be to avoid any cruise near the British Islands, or rather change your name to Hugh Williams if you want to be on the safe side.

This type of astonishing coincidences coined the concept of serendipity, first introduced by Horace Walpole, after the Persian fairy tale The Three Princes of Serendip, whose heroes were always making discoveries -by accidents and sagacity- of things they were not in quest of. As with all serendipities, apparently all three accidents in our story were unrelated, but maybe there was a hidden relation that made the outcome of them unavoidable.

Sometimes we wonder if the world and its laws create a fixed destiny. In a deterministic scenario, by knowing all the inputs and context conditions the outcome of any given situation could be accurately predicted. That could well be the case, unfortunately it would require god-like insight and unlimited time to know all the variables that can affect the final results of a certain action in a specific moment. That’s how probability is introduced in our interpretation of reality. Based on a limited background knowledge, and on our previous experiences, we can predict the likelihood of certain results, which usually is good enough to make informed decisions.

As technology has evolved over the decades, our calculation capabilities and the amount of data we are able to process has increased dramatically. And this trend is bound to continue in the future with the advent of new promising innovations, like the quantum computers and new materials like graphene, that could multiply by several orders of magnitude our present computers’ performance.

Additionally, we are generating an unprecedented amount of data, coming from millions of devices, sensors and their interactions, as well as from electronic process transactions, allowing governments and companies alike to access information that was simply not available before.

So, on the one hand we are accessing more information, and on the other hand we have more processing power that we ever had before to deal with it. These two topics pose their own unique technological challenges, more or less summarized under the Big Data umbrella, but apparently they get us closer to a deterministic characterization of the reality.

But are we getting any closer to being gods, to have the superhuman ability of predicting the future? Certainly we have more information, and the means to store it and process it in several different ways and faster than any time in our recent history, but this is just a part of the tale, and not the trickiest one. The real challenge is being able to make any sense out of it.

It is often said that information is power, which is only partially true. It is what we do with it and our ability to make the right predictions and take the appropriate actions from where power stems. Going back to our introductory story, it would have been impressive to know all the ships’ names in the world, their routes and the costs of all tickets, but for all the poor fellows in the last episode, the key thing they needed to know was a distinctive pattern that announced that if they were cruising those days around the British Islands, and not named Hugh Williams, they would have better remained on land.

So, do data tell stories? Are all correlations meaningful, or most of them are just misleading? As patterns are discovered, what is the value and impact of events that apparently defy such patterns? These are the tough questions that have to be answered in this era of information deluge, and going beyond Big Data, thinking about data interpretation through patterns and exceptions is the probably the right approach to this challenge.

Image source: Flickr.com


Categories: Scientific Community, Strategy & Innovation, Technology, Trends
Jordi Safont Guillen

Jordi has more than 9 years of experience in IT consulting. His assignments have included a wide area of topics from technical reviews to business process analysis. Within Atos, he has one of the first lean skippers in the transformation initiative for Service Desks and also one of the founding members of the Scientific Community.Jordi is passionate about learning with a practical approach, and committed to continuous improvement and excellence in business environment, with an special interest on how to make IT efficient and value adding for business through adaptation of lean techniques and focused innovation.Although based in Spain, Jordi enjoys working with international teams and having the opportunity to share his views and opinions with colleagues with different professional backgrounds and cultures.

6 Comments »

  1. Nicolas Roux says:

    Excellent article Jordi. Still, one the key challenge I see for Pattern Based Strategies is that correlation is not causality.

  2. bruno paul says:

    The answer is data interpretation but not exactly through patterns and exceptions. It is by modeling first our understanding of what reality is, then by interpreting data through these models, looking for the patterns and exceptions. Mining data, you can only discover new things you are indeed looking for. If we cannot guess what a correlation mean, it is useless in predicting any future. It becomes useful only when we understand the model behind, which explains the pattern. But many statistical correlations have no meaning at all. This means we cannot know better the reality, or the future, only by mining big data. There is a mandatory phase before aimed at thinking/creating the conception, and this phase is much harder.
    This is what European philosophy teach us, namely with Leibnitz and Granger.

  3. Aljosa Pasic says:

    I wonder if any Hugh Williams ever died in a shipwreck? Drawing patterns is fine , but we should be carefull about “black swan” type of events : in risk assessments there is a new term HILF High Impact Low Frequency events, which represents an important research challenge.

    • Jordi Safont Guillen says:

      Very good point Aljosa. In fact Pattern base strategy must deal as well with unexpected events which may be as important as those that were predicted.

  4. Rob Price says:

    Either Hugh Williams quickly became the name of choice for all Welsh & Scottish sailors based on his incredible luck in surviving the first, or he is a time traveller from the future who enjoys visiting ship wrecks from the past, or he is the one and only Highlander. Seriously, the co-incidence is uncanny, and i’d be fascinated to understand whether a deeper analysis of other informtation trawled from history actually finds a sequence of other Hugh Williams!!

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>