It feels like not a week goes by that a new headline suggests that AI has gone too far and we need to stop it’s advance before it’s too late. In a recent BBC TV documentary, Microsoft’s Brad Smith claimed we are moving closer to the fictional 1984. The big sound-bite from the documentary was that the dystopian future and the big brother surveillance depicted by George Orwell’s novel of the same name, could become a reality by 2024. It certainly is a headline grabbing statement and one I very much hope we aren’t going to realise.
AI is arguably one of the most important technological innovations impacting our society. But as it becomes more intertwined into our daily lives, it is coming under increasing scrutiny by law makers, who are seeking to regulate the technology. But should that be at any cost?
A recent Future Proof study found that the malicious use of AI could be the second most extreme risk after biological warfare and more dangerous to society than the current pandemic. Unsurprisingly, EU and US Policy makers are already looking at the real world implications of AI and are arguing for the need of tighter regulation. But opponents of such regulation argue that this could limit entrepreneurship and innovation and hand the advantage to China.
There are of course many examples of AI being used to the benefit of society. Cancer diagnosis, weather forecasting, better food production, climate change, wildlife conservation and even identifying fake news. Although the latter is a much bigger issue that goes beyond the implications of AI.
Brad Smith’s observation is of course a valid one. We only have to look at the vast surveillance apparatus that are now in place in many major cities around the world. Wired recently reported that Amnesty International had identified 15,000 surveillance cameras in 3 of New York’s 5 boroughs. The cameras are used to identify criminals but critics warn it increases mis-identification and that many of the cameras are clustered in non-white neighbourhoods.
The military are also experimenting with AI, handing off control to machines in complex, fast moving situations. But whilst the thought of this does evoke doomsday terminator like machines, researchers have proved that autonomous machines outperform humans when placed in high speed, high pressure situations.
And then there are countless examples of less intrusive and more novel examples of the application of AI. In sport, Deepmind is working with Liverpool FC to help players make better decisions, providing analysis on 1,000’s of passes, set plays and goals. Or there is the AI being used to produce poetry, music and even art. And if you are a Game of Thrones fan eagerly waiting on George R.R. Martin finishing the next instalment of the popular series, you’ll be pleased to hear that AI has begun writing the beginning of the 6th book. Winter is therefore kind of on it’s way!
The arguments for the machines taking over are of course, in theory, valid. But until it can provide 100% accurate and error free predictive text on my emails and messaging or serve me up exactly what I asked for on Alexa, I’ll take my chances and believe that the machines won’t take over anytime soon, at least not before 2024.