Artificial Intelligence is one of the big topics at the moment. All of the major IT companies are putting a lot of effort into this area. Both Google and Facebook recently announced AI as a core part of their development plans and Microsoft actually released some of its artificial intelligence projects into the wild.
So What Is Artificial Intelligence Used For?
When we think of artificial intelligence we usually imagine a robot.
But the artificial intelligence that’s currently emerging is much more subtle.
Yes, we’ve got things like the Google self driving car and a range of human and animal like robots that can perform various tasks, but it’s the artificial intelligence software that’s starting to wriggle its way into our everyday lives.
Many large corporations now use computerised customer service bots. These bots can understand human speech in a range of languages and to some extent can work out what the customer wants and then decide how best to deal with that enquiry. This allows them to filter calls so that only those that need human intervention get passed to a real operator.
Share trading has seen a massive increase in artificial intelligence over the past few years. Computer systems monitor the markets and use learning algorithms to predict share movements. Many are directly connected to the trading systems and will actually buy and sell shares without any human intervention. The notion off pits full of men wearing red braces and shouting and waving bits of paper each other is becoming part of our history.
Get Ready for the Bots
AI bots are the new playthings for the major IT companies. These are pieces of software designed to help you carry out a range of tasks.
For example, Siri on your iPhone and Cortana in Windows are both bots. You can talk to them, ask them questions or give them instructions and they will find information for you or perform various tasks on your computer.
But these bots are now being expanded into other areas. Facebook has launched a range of bots in its messenger app. These are designed to help and advise people by looking at what they are doing and what they are saying. Other bots will help you find and buy products by simply talking to your computer, book a taxi ride, send flowers, etc. More and more of our interactions with computers and the web are being monitored by bots who will influence what we see and the choices available to us.
How Intelligent Is Artificial Intelligence?
Artificial intelligence is still in its infancy. It can do some specific tasks very well, better than their human counterparts. For example finally beating the human Go champion. But they still lack the full understanding of the world we live in and implications outside of their area of expertise.
Even though the latest learning algorithms are giving the artificial intelligence software a deeper and deeper understanding of topics, they still rely on human intervention to provide guidance and understanding.
How Accurate Is Artificial Intelligence?
AI software is still being written by humans. Once it’s running it can learn and make amazing deductions from the data it gathers, but it can still only operate within the parameters set by the programmers.
Sometimes this can go drastically wrong.
In 2012 a major US financial firm, Knight Capital, had a software glitch in its trading system. It suddenly decided to start buying high and selling low. It took the company 30 minutes to notice the error but by that time the system had lost over £250 million. The software is AI engine just didn’t understand that the decisions it was making were wrong.
More recently Microsoft launched a chat bot called Tay. It was designed to behave like a young adult (18 to 24-year-old) and interact with people of a similar age on Twitter. The idea was for Tay to chat with people and learn from the discussions it had. It would then build a working knowledge of people’s interests and opinions so that it could have better conversations with them.
Unfortunately Tay had to be taken off-line after only one day. As you might have guessed when teenagers learned that what they told her would influence the way she thought, they started brainwashing her with a range of unsavoury opinions. This turned her into a radical right-wing, white supremacist, racist, antifeminist, Nazi sympathising embarrassment for Microsoft.
Where Is This Going?
Mark Zuckerberg has said that he believes artificial intelligence will outperform humans in a range of areas by the mid-2020s. It still won’t be conscious of the world around it or have imagination or be capable of spontaneous thought, but it will be able to see better, understanding what it sees, hear better, making sense of what it hears, and converse with people as fluently as a real human.
As humans we do need to be aware of this move. Bots will start to take over a range of jobs by being more efficient, more accurate, working 24 hours a day and costing only a few pounds worth of electricity per month. Initially these will be the more mundane jobs but as the technology advances, and it will, it’s hard to work out which areas the bots won’t be able to infiltrate.