Forecasting disruptions

I came across an interesting video of a lecture by Toni Seba, of Stanford University, in which he explains his theory of what he calls ‘technology disruption’. He looks particularly at the historical failures of corporations to predict the rapid rate at which one technology will grow and replace another – digital photography replacing film, for example. He suggests that we are now only a few years away from the total replacement of individually-owned and driven internal combustion engine cars with huge fleet-owned driverless, and cheap, electric vehicles.

Several things about this struck me.

Understanding unforeseen technology disruption seems very much like trying to understand and predict accidents. In safety we focus on unforeseen, usually adverse, events, and in Toni Seba’s case he focuses on unforeseen beneficial events. But there are parallels.

In safety we have theories about accidents which we study using tools like Events and Conditional Factors Analysis. We then use our findings to inform our approaches to risk assessment. It is impossible to foresee the future in detail, but in safety (as in any endeavour) we try to predict and take the right measures to ensure success and the avoidance of harm/loss. Over time we get better at it but bad things that we have not foreseen still happen.

As with an accident, in order to have technology disruption you have to have a successful concatenation of events and conditions. A lot of things have to come together at the right time and in the right way. So presumably there are a lot of potentially useful technology disruptions that might have happened but didn’t quite make it. Like ‘near miss’ accidents, we never really hear about them.

But history shows you cannot afford to be complacent, because change will always happen and in ways we cannot entirely foresee.

New technology emerges from a mass of apparently chaotic and highly energetic effort by a number of highly creative people. For it to succeed and be widely adopted, as with accidents, suitable and sufficient conditions have to be in place; you need the right triggers, and you must not have any significant barriers in the trajectory to success.

In its detail it is highly complex and after it has happened you can never uncover in total detail exactly what happened, when, where, how and why. All histories are approximations. You cannot rewind the great video of time and go back. It has happened and it’s gone. All you have is the evidence (partial memory?) left behind.

What Toni Seba seems to be doing is trying to deduce some universal laws about technology disruption based on costs and efficiency and what is known about past market behaviour – but as we know, forecasts are notoriously incomplete. Investigations tend to show that it’s the small details about unforeseen events which make all the difference. So, being incomplete in the details, all forward looks are approximations. But you should never stop trying to improve your forecasting.

Toni seems to be encouraging businesses to get much better at that when it comes to technology.

When it comes to mass transport based on fleet-owned, totally autonomous vehicles, his case sounds compelling and logical. Everything could be cleaner, less congested, cheaper and safer too. One hopes there are not going to be too many glitches. But other things might conceivably come into the equation such as systems security, insufficient progress in further battery, development, resistance to loss of social status associated with private car ownership, people not wanting or needing to move around so much, breakdown in the administration of city life etc.

So, how do you model all that?

Personally I think it pays to always believe that the future is bright, but we all know that clouds exist and that there might be a big one on the horizon any time soon. That’s a thought no safety conscious person can afford to ignore.

Roger Bibbings, partnership consultant