When Will Artificial Intelligence Change the World?
Are we at the onset of a world dominated by intelligent machines? With the greatest thinkers of our generation, from Stephen Hawking to Elon Musk, queuing up to warn us of humanity’s impending destruction at the hands of intelligent technology, should we be worried?
The rise of Artificial Intelligence (AI) in robotics is not as farfetched as it may seem. From automated farmhands to driverless cars, we are seeing a rise in the number of technologies that are able to perceive their environment and make a decision on how best to proceed in the completion of their assigned tasks. Robotic intelligence is growing at a steady rate; AI has now passed self-awareness tests it was thought only possible for humans to handle while a computer programme developed by Google’s DeepMind has recently beaten the world’s best player in a game of Go.
And even if we don’t follow the alarmistic point of view, AI has prompted a number of deeper questions: is it really useful? What sort of future can it offer businesses and wider society? Can a machine compete with humans at an intellectual level? Can a machine become dangerous?
Before answering any of these questions, let’s first try to understand what Artificial Intelligence really is.
Defining Real Intelligence
We must first make a distinction between robots and AI. While robots are simply the mechanical devices, AI is the ‘spark’ that helps them operate. But what is intelligence?
Critically, it seems that the ability to learn could be the differentiator. Intelligence is not simply carrying out a task but learning when, where and how that task should be carried out. This is what makes the aforementioned DeepMind AlphaGo project so exciting: although the machine was programmed with the rules of Go, it eventually learned to master the game by playing thousands of matches against humans (as well as millions of games against itself).
This is why programmes such as Siri and Cortana do not qualify as real intelligence. As human as they may seem, they are simply an automated search tool, albeit an impressive, voice-activated one. They can only answer one question and are unable to hold a conversation in its truest sense.
Bringing AI into the Real World – How and Why is it Used?
AI is now used in most applications that require analysis of huge amounts of data; and weather prediction, oil exploration, language and graphics recognition engines are some of those applications. AI can augment many job roles – allowing for intelligent decisions to be made in areas such as predictive maintenance for telecoms or manufacturing businesses. AI can help deploy staff more efficiently; ensuring that employees have all the necessary information needed to do their jobs properly.
Coupled with sophisticated language recognition tools, it’s not hard to imagine a world in which IT support or customer complaints are handled by machines. Of course, this throws up serious questions about the right application of AI – would you feel comfortable describing a problem to a machine? Although it may be able to provide a solution, AI certainly can’t offer empathy or understanding.
And what happens when AI goes wrong? The controversy surrounding Tay, an AI chatbot created by Microsoft, shows that the technology is still vulnerable to abuse. Within only 24 hours the organization was forced to deactivate Tay after a “coordinated attack by a subset of people” targeted the chatbot’s ability to learn from its interactions, feeding the programme offensive and racist information.
Ultimately, AI, when used correctly, could have a huge benefit for both businesses and wider society. Taking on the simple and repetitive tasks will help further push our transition to a knowledge-based economy. There will be change, and although AI is advancing quickly, there will be time for us to adapt to this new world.