Peter Cochrane argues that our world is essentially non-linear and becoming more complex. This complexity is visibly pervasive in our workplace, government and everyday living. Furthermore the thinking and models of the past industrial revolution and experiences of past centuries cannot solve today’s problems as we try to engineer a future that supports a growing population without destroying our environment. He discusses how Artificial Intelligence may form part of the answer.
Like all school, college and university students I was fed a diet of simple, linear and well-behaved problems with easy-to-define solutions. And whilst this academic convenience made it easy for teachers and professors to set exam problems and grade and assess the performance of their students, I later found it to be a far cry from the real world experiences of industry. Throughout my professional life those initial ‘innocent simplicity’
- assumptions of linearity may have started as a point of nucleation for my early thoughts, but they have inevitably been cast aside at an early stage! How come? The universe and everything in it, including our networks and working environment, is essentially non-linear and complex, and for the human race it is becoming ever more so with each new layer of technology and the insights that it affords.
Sources of complexity
Where does this complexity come from? Well, it really is all about networks, connectedness and delay. Quite remarkably, and until very recently, all human progress was founded on simplistic models of the physical world, technology, society and peoples. And this was really successful because communication and interaction were disjointed and slow, and it was possible to get by with simple-to-understand models and basic decision trees involving a small number of parameters. But because the non-linear effects occurred slowly the perception was one of well-behaved linearity. Of course there were a few exceptions that were not fully understood at the time such as the financial collapse of the 1930s, Eastern Seaboard (USA) Power Brownouts during the 1970s, telephone network voting, which suddenly imposed massive traffic demands, and of course a raft of military and political decisions.
Unfortunately, the growth of connectedness combined with the removal of delay increasingly confounds the simple world-view, and it is now clear that the occurrence of linearity is the (rare) exception. Take banking and finance: I think we can safely say that recent global events have shown that no one understands how it all works. The rapid rise of chaos (in the mathematical sense) that crashed the world economy can be assigned to greed, a lack of understanding, high-speed networks (connectivity) and the reduction of transaction delay driven by competition. In effect, it was as if the entire system was riding a bicycle and got into a speed wobble, and the rider was unable to get back to stability without having an accident!
On another plane of simplicity; the proponents of genomic decoding thought that individual genes would be responsible for individual effects. How very unlikely! It not only turns out that combinations of genes are responsible for given effects, such as defective human organs and disease, but this is also influenced by the function of proteins. This is far more complex than anyone first imagined and will take considerable computing power to unravel. One supposition that does seem likely, but has yet to be proven, is that cancer is merely a communication problem between protein and genome rather than something induced by food or our unhealthy habits. The snag is that proteins are about 1,000-fold more complex than the genome, and the computational effort will therefore be far greater. The good news is that Gordon Moore’s Law will see silicon computing power continue to double and take care of the shortfall in a very short time!
Complexity is now visibly pervasive in everything from our IT to the workplace, government and everyday living. Global warming, recycling, energy, food, water, logistics, sustainable manufacturing and most other important issues are not simple, isolated and bounded problems. The thinking and models of the past industrial revolution and experiences of past centuries cannot solve these problems as we try to engineer a future that will support a growing population without destroying our environment.
There are lots of simple solutions to complex problems – unfortunately they are all wrong!
Not surprisingly then managers and politicians are struggling to make wise decisions. Their world-view is from a past era of infinite supply, linearity, disconnectedness and continual growth. If we are to engineer a world that is sustainable we have to think afresh and move away from the simplistic management of growth, profit and return on capital to something that encompasses the welfare of everyone and everything. This will involve the rethinking of industry and supply, our logistic systems, mode of trade, and the way we value commodities and increasingly rare materials.
Money alone is far too crude a measure to make any decision and we need to at least account for Ecological Impact and Social Consequence
To survive and prosper in the long term we have to embrace new opportunities that span the worlds of the organic, inorganic and technology, with base solutions that are holistic and resilient, and our core mission from now on has to be:
More from Less for More and More People
More from Less for The Few
How do we achieve this? First; we only have one visible route to sustainability, which is at the intersection of IT, nano and bio-tech. The ability to manipulate materials at a molecular and atomic level, and to programme their behaviours, opens up new avenues that use minimal energy and achieve nearly 100% reuse of materials. At a simple level the first car was printed last November, and we can now 3D print adjustable spanners and gearboxes in one pass. No multiple parts to be assembled, but a single integrated process.
To date our biggest successes are printable and programmable plastics, but we are making inroads into metals. So there is an expectation that production will be simplified and products distributed in ways that involve a much lower logistics cost. On an entirely different level, organic computer screens, devices and batteries have also been produced along with sensors more capable than a ‘dog’s nose’. An ability to detect a single molecule of a substance in a sample will transform medicine, food production, farming and the fast-growing care industry.
What was science fiction will soon become science fact
For example; all the components to make a ‘Tri-Corder’ (aka Start Trek – the TV series) are available, and in principle, the task is now one of ‘mere engineering’. So we might anticipate some very big changes in society and industry with new understandings and capabilities:
Machines smarter than people in specific sectors, more mobility and sensors, everything and everyone online, greater individualisation, new networks and networking modes, more accurate modelling and decision support.
The continued expansion of the internet, computing power and mobility is about to undergo another big change with The Cloud. Computers will collaborate on line, and the processing power of mobile phones, automobiles et al, including your TV and hi-fi, will be on the same network. Most importantly, this new ‘society of things’ will see sensors that will increase the information flow and intelligence with significant benefits for everything from integrated transportation to logistics, retail, and human creativity and problem-solving.
Today we are constantly surprised by Artificial Intelligence (AI) systems and the answers they contrive, and on many occasions we lack the facility to fully understand. But that does not preclude us using the results. For example; IBM with Deep Blue, Blue Gene and Watson presented us with the first public surprises, but Siri on your iPhone 5 is now probably the most public, whilst Watson is being used by the medical profession in the USA. Almost by osmosis we have gradually realised that the solution of many industrial, scientific and governmental problems will continue to defy human abilities. So the question is; could AI provide further and significant enhancements to the world of business and problem-solving?
Undoubtedly, the answer has to be yes! But machine intelligence has been around for decades and we have to ask; why don’t machines understand? The answer is simply because we have only recently assembled sufficient data and a number of ‘common sense cases’ to afford them some context to the human situation. Also, the computational power required for cognition has not been available – but now it is! More importantly, AI systems are now capable of learning from a vast set of real-time inputs in the form of text, audio, video, and animations and simulations, plus the impending roll-out of dedicated sensors embedded in billions of mobile phones. On a professional level; AI systems provide far better medical diagnoses and prognoses than a human doctor, and systems are now being rolled out along with all the legal trappings!
The lack of sensory devices to give computers continual input and some form of self-awareness has been a misunderstood subtlety
So where is AI today? The big successes go largely unnoticed and include control systems in industry and aviation, financial trading systems and telecom network control, the design of integrated circuits, and complex industrial process and plants. AI systems also populate many control systems for our cars, elevators, trains and domestic appliances. On the more esoteric side AI systems now make discoveries in science and invent/innovate in technology and engineering.
Science fiction writers mostly predict Armageddon and the machines taking over, pushing us aside and becoming malevolent. My prediction is that we will be happy to let go of a lot of things, like mundane processes and production work, and form a partnership with machines so we can embrace the more creative and productive aspects of work. At least that has been the history so far!
All of this prompts one key question: will we be smart enough to recognise new intelligences when they spontaneously erupt on the internet or within some other complex system?
Of course, there is another question that you might want to ask as a manager and/or a player in business, government or some institution: what does this all mean for me, my work and my organisation? It depends on where you are in the organisation and your sector. In the military and defence industries people are already enmeshed in and trying to deal with mounting complexity, whilst charities seem to be largely unaware. So my advice is to scan other industries to see what they are encountering, read, attend conferences, look out for the latest tools, and start thinking about business models, modelling and applications.
At the two ends of the spectrum we have the producers of commodities; concrete, steel, cars and food et al where everything about the process is known and the variability of demand and specification is always on the move. Here, hierarchy and clear job demarcation works! Here chaos comes in the guise of weather, fashion, seasonal demand and rabid competition.
In complete contrast, and at the other end of the spectrum, we have R&D, design and consultancy where even defining the problem can be difficult and might not be stable! And hierarchy would certainly be the kiss of death. Agile, free thinking, multi-role and multi-talented people/organisations are the name of the game here with customers often satisfied with any kind of a solution.
And so the most complex organisations embrace the complete spectrum and the managers need the wisdom to recognise that ‘one size does not fit all’ and each department has to have the freedom to operate in the mode that is best for it to achieve its objectives.
If the suits (managers) and sandals (researchers/creatives) can’t mix, then the result is usually divorce.
For the vast majority of organisations it pays to remain fluid, prepared to change and adapt, and always be ready to adopt the next wave of technology and management solutions. An acceptance that modelling and AI will be an increasing element of that change is also an essential. And if you don’t you can be assured that your competitors will!
I welcome your thoughts.