Post-Deep Learning: Is Deep Leading the Next Step?

Post-Deep Learning: Is Deep Leading the Next Step?

Published 11th August 2016
Dominique Sciamma

Managing Director & Dean, Strate School of Design

Published 11th August 2016

During spring 2016, AlphaGo, artificial intelligence developed by Google, defeated two human Go champions, thus breaking a frontier many people and specialists thought would never be possible. Deep learning refers to an artificial neural network that is multi-layered.

Based on neural networks and a quite old algorithm known as deep learning, this artificial intelligence has been able to master the most complex strategy game ever created, where the combinatorial explosion seemed to forbid any winning strategy based on sole computation. Go players and specialists pretended so far that only a human mind, if not a human spirit, could have the qualities, the intuition, the creativity to play such a game, and more than that, to win.

AlphaGo’s definite victory killed this prejudice, and seems to open a new Pandora’s box-- creativity no longer exclusively belongs to mankind, and hence machines and automation can potentially dominate other aspects of the human sphere.

If this is true, there is in front of us an essential and existential question: would machines also take over our capabilities to lead?

After deep learning, is deep leading the next step? If yes, does this mean that all humans projects – either big or small, industrial or political - and our individual and collective destinies, will be led by machines?

It is a very important question.

We are living a time where the promises of automation and robotisation, dreamt by science fiction authors and prospectivists since the beginning of the 21th century, are feared and wanted at the same time. While we destroy jobs, we also gain access to comfort and new objects and services. If this schizophrenic situation is the price and condition of the affluent society, we are more or less accepting it because we consider the balance to be positive.

With deep leading, it is not our mechanical capabilities that we expect to leave to machines; it is our capacity to analyse, to create and to make decisions. This is something right in front of us that we need to consider carefully.

Strange days indeed.

In an era where leadership has been identified as the concept, process, quality, behaviour that we should invest in, by training leaders through a reinvented education and for the benefit of all, AlphaGo is trumpeting that this is not an issue anymore and that machines will take over.

But no.

The very definition of humanity is that we transform our knowledge into technology, tools, theories, in an iterative way, to ease our single and collective lives, to extend our body and mind— not to make us less humans—but more humane. Computers, programmes, robots, AI are all the latest tools, technologies and theories, with an incredible power, to make us more human, more protected, freer.

Deep learning algorithms deal with repetition and combinatorial situations rather than with complexity. Dealing and playing with complexity demands a capability to represent, to invent, to think. Not the capability to recognise, not to consolidate, not to compute.

The good news is that AI will not take over our capability to lead, but it will contribute to leadership.

As ever, the decision is in our hands.

Back to top