John McCarthy, one of the fathers of Artificial Intelligence (AI), defines it as “the science and engineering of making intelligent machines, especially intelligent computer programs1“.
Put more succinctly, it’s creating machines that are intelligent, that do not only react to specific stimulus but can learn on their own. People often think of talking robots or the terminator movies but in reality, AI is much more diverse.
Simple or narrow AI is evident in many technologies today, the most popular of which is Siri by Apple or Alexa by Amazon. These AI systems are designed to learn our habits and mannerisms and respond to our requests; Amazon Echo, powered by Alexa will shop for you, schedule appointments and even allow you to control your smart house.
AI behind the wheel
Autonomous vehicles are another AI-enabled piece of technology that will have a profound global impact. It’s possible that many will have no need of owning a car, hailing a driverless vehicle from their smart phone from a pool of public cars. Alternatively, your own car could find its way home after dropping you to work, or pick the kids up from school, or even pick up your groceries via drive through supermarkets.
Regardless, licenses will be a thing of the past, accidents would drop dramatically, panel beaters, insurers and car dealers would struggle; entire city landscapes would change as car parks disappear and road networks become smaller and more efficient. Truck and cab drivers would no longer be required and in extreme examples, neither would pilots or ship captains! We would be able to work or relax in vehicles, optimising our time however we see fit.
“Watson” is IBM’s AI software that made its rise to stardom by winning Jeopardy! in 2011. Since then its capabilities have grown materially; it now has oncology applications which IBM claim that in some cases Watson can diagnose more accurately than human doctors. This is based on its ability to read medical evidence at a rate of 800 million pages per second, with access to over 23 million global research studies2.
To put those numbers in perspective, it would take the average reader 18 years to do what Watson does in a second! It is estimated that it would require at least 160 hours of reading just to keep up with new medical information as it’s published3 it would take Watson 0.001 seconds
If this software can eventually diagnose more accurately than most general practitioners, what happens to the health industry? People could be scanned from home with a virtual doctor assisting them, with the appropriate medicine delivered right to their door.
Watson is now turning its hand to law under the pseudonym “Ross”, pitched as an assistant tool for legal professionals rather than a replacement. However, if it can accurately advise on general enquiries in a matter of minutes or even seconds, what does this mean for paralegals and junior attorneys? It would drive costs down and make legal services available to the masses.
Can we control AI?
The interesting phenomenon surrounding AI is that if we allow machines to learn, they can very quickly go on a path none of us could’ve anticipated.
This unpredictability was evidenced recently when Facebook allowed two of their chatbots to negotiate with each other in English. The bots quickly realised this was inefficient and created their own language that was nonsensical to humans. Facebook had to shut down the program.
Another example is when live-streaming service Twix had two Google Home devices (similar to Amazon’s Echo) debate for hours on end. Some of the snippets included: “it would be better if there were fewer people on this planet” to which the other device replies “let it send this world back into the abyss”. They also eerily argued over which one of them was a robot and which was human, both believing they were the human. To be fair I’m paraphrasing materially, as they did go on to say how much they loved each other!
Fundamental changes coming
People debate whether we are even capable of creating a machine that thinks like a human, or if we should even try. However, the creation of a human-like robot to me isn’t what’s important- it’s the massive disruption that occurs along the way; impacting everything from global industries, to city landscapes, to shopping centres through to our lifestyle. These changes will fundamentally alter the way we live and interact with the world.