There’s no denying the hype surrounding artificial intelligence or AI. Most weeks will feature a media story questioning whether machines are set to take over the world and give rise to another industrial revolution.
Cambridge University recently announced the launch of a £10-million research centre to address some of the growing concerns over the impact AI will have on everything from the loss of jobs to humanity’s very existence. But just because a computer – admittedly a very smart one called Watson – can beat two quiz champions at a game of Jeopardy! doesn’t mean we should pull the plug on the progress that’s being made in machine-learning technology.
We are undoubtedly moving towards a symbiosis of man and machine, but a symbiosis where the human aspect is equally central to success
We are undoubtedly entering a new industrial age and will face some readjustments as a result, but not changes we necessarily need to fear. We have been through this hype cycle before in decision sciences and big data analytics. In the early-90s everyone was talking about data mining as the next big thing, and a lot of algorithms were built to try and support machine-learning. But those efforts largely fell short because all but a handful of organisations lacked the all-important ingredient of computational scale.
What is exciting now is not so much the work being done around AI, but that we finally have the full complement of ingredients in place for this computational environment to thrive. The information age has finally found its “steam engine” and is driving us forward into a new era, an age not just of information or intelligence, but organisational consciousness.
In practice, what this scale and technological alignment in analytics has enabled us to do is create algorithms that can support advanced machine-learning techniques which require a lot of training, a lot of data, a lot of iteration and a lot of computational power. It means we are now at a stage where we can finally churn out useful forms of AI and we should celebrate that. Think of the sensory-fusion deep-learning research being done by the team behind Brain4Cars, making cars safer to drive, and those organisations that are using it in the battle to cure cancer.
Does this mean that machines are getting ever closer to putting us humans out of a job? Are they now capable of handling the heuristic and algorithmic methods needed to solve complex business problems? No.
While science and machine-learning will always remain central tenets of analytics – where the science is expressed through algorithms and mathematical thinking – solving increasingly complex business problems requires a whole other perspective. Building intelligent systems, or what we at Mu Sigma call learning systems, requires the harmonisation of science and art, intelligence and consciousness, scale and diversity; a creative thinking process that contextualises design and behavioural understanding, which then become inputs to the algorithms.
The emergence of some of the headline-grabbing AI we’ve seen in recent times has to a large extent relied on pattern recognition to automate thinking and works in an exploitative mode. By contrast, we are now seeing the rise of a new AI – augmented intelligence – which works in an exploratory mode, and involves probing and asking the right kind of questions that enable new learning to emerge.
This first principles-based thinking sits at the very heart of the human psyche, one that always demands more and doesn’t like sitting still. In doing so we as a species constantly change our behaviours, and create evermore complex and interconnected problems for organisations to solve. And, because of this rapidly changing evolution of both needs and problems, along with their interconnected nature, organisations face the constant battle of needing more intelligence to throw into this problem-solving mix – and that all-important ingredient can only be provided by human intervention.
It’s hard to argue against the predictions that Ray Kurzweil outlined in his Six Epochs of Technology Evolution. Right now we are undoubtedly moving towards a symbiosis of man and machine, but a symbiosis where the human aspect is equally central to success. For augmented intelligence to be successful it requires machines to automate the science and provide the scale, and human beings to facilitate the art as well as the diversity of problems being solved. Only then can we create organic systems that do actually get better with usage and with more data, and create the kinds of solutions which will actually make a difference to the world.
I leave you though with the thoughts of Professor Huw Price, the philosopher who is directing the new ethics research facility at Cambridge University: “As a species, we need a successful transition to an era in which we share the planet with high-level, non-biological intelligence. We don’t know how far away that is, but we can be pretty confident that it’s in our future. Our challenge is to make sure that goes well.”
For further information please visit www.mu-sigma.com