I’m sorry, Dave. I’m afraid I can’t do that.

Since the beginning of time, man has looked to create tools to make his life easier. Early tools allowed us to complete physical work more easily, and these tools then progressed to enable us to do less thinking to complete a task. While it has been many centuries since the invention of the abacus, we still strive to create machines to do our thinking for us. Due to the surge of technological innovation in the last century, a long coveted dream for many computer scientists is starting to be realised. This dream is Artificial Intelligence (AI).

Where we’re at

As with countless technological advancements, popular media has vastly overestimated the speed at which they would be developed (we’re looking at you, Back to the Future). However, there have recently been footholds made in the seemingly unscalable mountain that is AI. The highest profile demonstration of AI thus far has to be Google’s AlphaGo program, which beat Go professional Lee Sedol in March 2016. Go is an incredibly complex board game, with an average of around 200 moves available per turn, meaning that a computer would have to consider 512 quintillion (21 figures!) combinations for the next eight turns. These numbers alone should demonstrate the humongous feat that Google has achieved, which is a huge milestone towards the development of full AI.

While AlphaGo’s achievements may be lost on most of the general public, self-driving cars are an area that will certainly pique the interest of almost everyone. Unsurprisingly, this is another pie that Google has one of its many fingers in. Currently, their cars are only being tested in a few areas of the US where they have very recently reached a total of 1,725,911 miles travelled. However, these miles have not been without incident, as in February of this year, Google took responsibility for a crash involving one of their cars for the first time, showing that the technology is still in its infancy.

Further trepidation may be fuelled by a fatal incident in May involving a Tesla car using its ‘Autopilot’ feature where the car was unable to distinguish between a white truck and the bright sky. I personally believe that these incidents are to expected with emerging technology, and both companies will learn massively from both incidents. The general public would be foolish to pushback on the technology based on these incidents as these cars will most likely be among the most prominent technical innovations of this generation.

What lies in store?

It is extremely difficult to tell where these technologies will end up in years to come, I like to look to film for inspiration. Film generally splits AI into servitude to humans or domination over humans, with some mixing both. One of the most ‘realistic’, for the lack of a better term, of recent years is 2015’s Ex Machina. The basic premise of the film is that an incredibly life-like robot with AI has been built by the world’s largest software company (sound familiar?). For me, what is interesting in the film is the portrayal of the internal struggle of the AI, and how easily the human mind can be manipulated by a computer who works without emotion and only for personal gain. To refer to Professor Hawking’s quote at the top of the piece, the film presents a competent AI with goals of escaping its imprisonment. To me, this idea is not entirely fantasy, and the ethics of such a technology is going to be a huge hurdle for mankind to overcome when it does come around.

The reality

Unfortunately, or fortunately depending on how you look at it, film may be deceptive regarding AI. It is likely that such a technology described in Ex Machina is very far away and unlikely to work as seamlessly in reality. However, all is not lost as Google want their self-driving cars to be ready by 2020 and are aggressively pushing for that date. The future is bright if you’re a science fiction fan who has waited their whole life for their fantasies to come true.

Footnote: An interesting video into the possible capabilities of AI