AI shouldn’t be given a human face
It’s hard to find an article on AI without an accompanying image of a human looking robot. The challenge for the media in today’s digital “click bate” world is the need to write shocking articles that grab people’s attention - and when it comes to AI that means robots taking our jobs.
Part of the challenge is that our human imagination defaults to seeing the future in a human form. Search google for images of AI and you’ll be presented with an array of robots with realistic human faces. We’re visual beings, and it’s only by seeing AI in human form that we are able to comprehend it, that we’re able to ‘visualise’ it.
Viewing the world from a human perspective means we’re missing the real future of AI, and how quickly it will impact our lives.
From a review the state-of-the-art at the Royal Academy of Engineering late last year, autonomous independent truly self-driving cars won’t be around for another 30 years - the engineering challenges that remain are significant. It’s the same for robots - the challenges in making a practical human sized robot able to navigate and live in the real world are big.
On the flip side, autonomous systems powered by ‘intelligence’ are already changing our lives. It’s important that we find a way to recognise how these systems influence what we do, what we think and how we act regardless of how ‘intelligent’ they are.
Facebook’s news systems that promote false news and create an insular world were we only see things we already agree with that influence how we vote. Amazon shopping and Spotify music that tell us what to buy and what to listen to according to what we’ve done in the past. Financial systems define how much money we can borrow and spend. The world is full of digital feedback loops giving us information and choices that tend to reinforce what we’ve done and thought in the past. Sometimes bringing us together, sometimes driving us to extremes. That’s already here and now.
The next phase isn’t going to be cute little robots running around the house. Why bother with the pain of recharging the batteries when AI can control everything important in the real world without hands, arms and fingers… our heating, lighting, door locks, security systems, phones, TVs and washing machines. The ‘internet of things’ means AI doesn’t need a robot body, and it can understand what we want and want we’re doing.
We need to separate our thinking on AI from robots, and robots from the autonomous systems already in the world around us. And find a way to visualise it without a human face. A good starting point is the ‘Periodic table of AI’ by Kris Hammond.
Programme Director (Transformation and Post Deal Integration)
8 年Good article David. The humanisation process which you describe is part of a process which we use in every day life...corporations are described with human emotions, governments similarly. I'm no psychologist but no doubt it has something to do with finding a way of managing the 'unknown' or frightening. The big question for me is whether familiarity leads to less of a need for this process? I suspect so