How true are hyped predictions that in future you could be asking an algorithm for a raise, instead of your office manager?
Dheeren Velu, who focuses on guiding organisations in adopting artificial intelligence and cognitive technologies, says the reality is a little more nuanced.
“If you look at it, traffic signals, they used to be managed by a police officer that directed traffic, but now we are so used to automated traffic lights,” Director of AI at French Consultancy, Capegemini says.
“We’ll see more of that, when it comes to allocating work,” he says.
“An algorithm can probably do that better – allocating work to the right people – than a traditional manager.”
AI no match for a child
General intelligence, however he explains is still very much the human’s domain.
“It’s when you can do a wide range of tasks.
“We will continue to excel in that whole interpersonal space, creativity, social interaction, leadership – situations which are unpredictable.
“As opposed to very specialised routines, things that you can draw rules and lines around, machines will definitely be much better than us at that. We will continue to excel in general intelligence.
“We have this unique ability to learn from a certain situation and apply that in another situation.
“My five year old son has more general intelligence than the smartest computer.”
A report into the effects of AI from the Australian Council of Learned Academies, estimates that 3.5 million Australian workers are at high-risk of being displaced by automation between 2015 and 2030.
But Velu says the relationship between humans and AI in the workforce is also one of collaboration.
“Generally as a guide line, there are certain things that we excel at and certain things that machines excel at,” he explains.
“The ability to achieve very specific goals to very specific solutions – we call that specialised intelligence,” he says.
“While the ability to achieve a wide range of goals in a wide range of scenarios, we call that general intelligence.”
“Machines are very good at specialised intelligence.”
As an example he uses a complex algorithm developed by Google’s DeepMind technologies.
“You take AlphaGo, one of the most sophisticated algorithms created.
“It plays the game Go, which is one of the most complex games ever and it won against the human champion, but if you asked that system to play a game of tic-tac-toe it would fail miserably.”