Friday, June 13, 2014

Robots will eat all the jobs ...

In 1896, N.Hawkins wrote the New Catechism of Electricity which included in it a warning on the impacts of electricity on jobs. In 1962, D. Michael wrote Cybernation : the Silent Conquest which included in it a warning on the impacts of computing on jobs.

Throughout history, at every age from industrial, mechanical, electricity, computing, internet and today then people have written warnings that this time, whatever it is will eat all the jobs. Everytime they've been wrong.

To understand why, I'm going to quickly summarise a cycle I've talked about for the last half dozen years - peace, war and wonder. A more detailed version can be found here

All our industries consist of value chains comprising of activities, practice, data and knowledge which are evolving due to supply and demand competition. For activities we describe this path of evolution as genesis, custom built, products (+rental services) and commodity (+utility services) and you can visualise the environment through the use of maps (see figure 1)

Figure 1 - Example Map from High Speed Rail overlaying different methods (due to changing characteristics)


Maps are also good for organisational learning and in detecting common patterns in the wider economy. There are many of these. One very important factor to understand is that components (whether activities, practices or data) change in properties as they evolve (see figure 2) due to competition i.e as an activity shifts from left to right on the above graph, its properties change.

Figure 2 - changing properties of evolving components.


This change of properties in combination with competition and inertia creates a common re-occuring pattern known as the peace, war and wonder cycle. The cycle appears both at a local industry level and macro-economic scale (the largest being known as K-waves or 'ages' and is directly attributable to how widespread the underlying components that are changing are in value chains across industries). The cycle itself mimics the adaptive renewal cycle that is found in other systems (see figure 3) and basically involves commoditisation of the past and the rapid creation of the new.

Hence (from figure 3) as an activity A evolves (due to supply and demand competition) through various states A[1] to [A6], it passes through a relatively peaceful state of competition between big vendors (A[1] to A[5]) that all build inertia to change due to past success. It's normally a company that is not encumbered by past models (think bookseller for infrastructure) that shifts the act from a product to commodity (and / or utility services) i.e. A[5] to A[6]. This change is rapid (known as a punctuated equilibrium) and is associated with changing practices (for computing think DevOps) and the rapid creation of higher order system e.g. B[1]. The rapid explosion of higher order systems that are inherently uncertain in value (think age of wonder e.g. utility provision of electricity enabling a mass of new activities), creates new forms of data (think Big Data associated with the shift of computing from product to utility) along with disruption of past models and new organisational forms. You can think of the stages as being equivalent to the adaptive renewal cycle with Peace being equivalent to exploitation and conservation, War being equivalent to release to re-organisation and Wonder equivalent to re-organisation to exploitation.

Figure 3 - Peace, War and Wonder cycle


The interesting part about this cycle, is that during the 'war' phase we get co-evolution of practice, new forms of organisation appearing and disruption of the past combined with a 'wonder' phase where new forms of activities and capabilities appear. This pattern isn't new, I used it when writing Canonical's play in the cloud in 2008. I also used the pattern to determine the new phenotype of organisation that was emerging in 2011 (see figure 4).

Figure 4 - Phenotype of new organisations forming in last cycle (caused by commoditisation of parts of IT)


The economic system turns out to be far more predictable than most people believe but then given most people can't even see the chess board they're playing on then it doesn't surprise me that most think it's random. 

Now, the key part of the peace, war and wonder cycle is that we can see the past that is being industrialised but we cannot see the future (the 'wonder') because it is by very nature uncertain, uncharted. 

Throughout history, the new activities and new capabilities have created new jobs e.g. electricity might have decimated the gas lamp lighters role in life but it created radio operators, TV presenters, computer operators and a host of other roles. However, those new activities are uncertain and hence we don't know what the impact will be.

We know that 'Robots will eat many old jobs' but can we say 'Robots will create many new jobs' ... well, that's the tricky bit. As Jim Stodgill correctly points out, it is a statement of belief either way. This is because of the uncharted (hence uncertain) nature of the new activities. We don't know, we cannot know and there is no crystal ball. The future is actually an uncertainty barrier for information.

Fortunately in the past it always has created many new jobs despite people writing that the change will destroy work. Unfortunately, given a long enough timeline then even Chicken Little will be right.
Post a Comment