Artificial Intelligence: How Can We Ensure This New Technology Serves Us?

The answer lies in understanding that today’s AIs are very narrow – they can do certain specific tasks, but only those tasks, astonishingly well.

Where next for humankind; the apex primate?

When you look at the mammal that is Homo sapiens in the context of geological time, the approximately 200,000 years we have been around is a tiny, tiny amount of time — less than 0.005 percent of Earth’s existence.

 And yet, so significant has been our species’ impact on the planet, that climatologists and geologists have named our era the Anthropocene epoch, because for the first time in the 4.1-billion-year history of life on Earth we humans, as a species, are changing what happens to and on the planet significantly, rather than simply being the observers and subjects of natural forces.

 In that 200,000 years since Homo sapiens first emerged in Africa and spread across the planet, we have evolved to become a “reasonably smart” apex primate at the top of the food chain in a closed system called planet Earth.

As we celebrate International Biological Diversity Day this month, it’s worth taking stock of where humankind finds itself, and what the future holds for the species at the top of the food chain.

The wonders of the human brain

Humankind’s “reasonable smarts” come courtesy of what might be the most amazingly complex thing in the universe: the human brain. An estimated 86- to 100-billion neurons and 7,000 times as many synapses account for roughly 2 percent of our body mass, yet consume around 20 percent of the oxygen and energy we take in.

It is our brain and its astonishing capacities that keeps us alive on daily basis, that enables us to dominate other animals, and that accounts for all human progress and the massive impact we have had on the planet despite, having been here for only a negligible amount of time in the grander scheme of things. It’s this amazing brain that has helped us to develop and master the various technologies that have brought us this far; from fire to fission and everything in between.

But it may not be sufficient to ensure our survival as a species. Our individual brains are ill-equipped to deal with either the size or dynamism of some of the challenges we now face.

We will need to look elsewhere for solutions to the many pressing problems, such as how to feed an additional 2-billion people on shrinking amounts of arable land, how to manage traffic congestion and safety in rapidly urbanising societies, how to provide sufficient energy for economic development, or manage epidemics, or maximise corporate profits without harming society, and so on.

Augmenting human capabilities

Help is at hand. We now stand at the dawn of a digital revolution; one which promises socioeconomic upheaval as profound as that which followed previous agricultural and industrial revolutions.

Whereas the plough, the steam-engine and the production technologies of yesteryear augmented our physical capabilities, this new digital revolution — with its data generation, information processing and communication technologies — is about augmenting our mental capabilities. Pre-eminent among the multifaceted technologies that underpin the digital revolution is artificial intelligence [AI].

As we embed sensors in more and more things in the world (including ourselves), and this cyber-physical world creates unprecedented volumes and velocities of data, we must use AI to make sense of it, as our brains are simply not up to the task of dealing with the velocity and volumes of data.

Unlike our brains, and courtesy of Moore’s law, we can scale up silicon-based capabilities in a largely unrestricted fashion — it’s not bound by the physical limitations of the human skull or by the metabolic need for sleep — and this is enabling us to deliver AI capabilities that were the stuff of science fiction only a few years ago.

There is much debate as to when (and whether) AI will exceed human intelligence, with futurists such as Ray Kurzweil (who claims an 86 percent accuracy rate for the 147 predictions he has made since the 1990s) positing 2045 as the year when “the singularity” will occur. “The singularity” is the point in time at which AI leads to machines that are smarter than human beings.

We’ve heard such predictions before, and they’ve proved largely false. But there are differences this time around that we should take heed of, if we are to apply AI to serve humanity and make the world a better place.

New technologies enable new possibilities

Improvements in computing technologies, especially graphics processing units (GPUs), have led to the point where supercomputing and massive memory is affordable and therefore widely available.

Significant improvements in algorithms (not least of which is deep learning) and massive sets of data on which to train new AI models using techniques like machine and deep learning is set to accelerate as the internet of things (IoT) takes hold. The possibilities are endless, and limited only by our imaginations and ethical considerations. The ramifications can be scary, and the process is unstoppable.

It is now up to us as individuals, workers, parents, managers, leaders, companies, governments and societies to understand and evaluate these trends and technologies and to ask: how can we ensure this new technology serves us?

The answer lies in understanding that today’s AIs are very narrow — they can do certain specific tasks, but only those tasks, astonishingly well. So it’s about picking the most valuable-use cases, understanding that for certain tasks, where efficiency, repeatability, neutrality, speed and big sets of data are the norm, the narrow intelligence of today’s AI is often superior to humans when it comes to getting a job done.

It’s also about having a mindset of embedding these new tools into both existing or new business processes and models in such a way that we can take the “work out of work” and free our people up to do the things that only humans can do; imagining, empathising, relating, creating and solving complex problems. This last point is worth emphasising; AI cannot envision, it cannot innovate, it doesn’t empathise, it cannot synthesise new solutions to complex problems.

It is about taking tools such as machine learning and applying them to the data you already have (or will generate through new capabilities such as social listening, IoT or visual processing) to generate new insights, to make life and work safer, easier and more productive, and to design innovative competitive capabilities.As we stand at the beginning of a new age for humanity, one where we can use AI for good, it is up to us to explore ways to make sure technology serves us well. We don’t yet know where it will take us, but we do know that we must get started.