“AI is just fancy maths” according to innovation specialist, digital ethicist and self-proclaimed ‘data philosopher’ Charles Radclyffe. He was speaking on a panel discussion that kicked off The Drum Big Bang conference, in partnership with Marin Software, at The Science Museum. “It’s fancy maths plus data plus computational power,” he added.
The dream of matching machine intelligence with that of humans is nothing new. However, the current hype around AI, arguably one of the most important technological advances of the 21st Century, has reached fever pitch. Are these hopes misplaced? When it comes to AI, are we in an era of over-optimism to be followed by a period of deep disillusionment? – these were some of the themes explored on the panel, moderated by associate editor, Sonoo Singh.
Reality vs Hype
“There is a huge difference between the potential and opportunity of automating human tasks – and automating the human itself,” said Sera Miller, co-founder of The Fawnbrake Collective. “I do get frustrated by [Hollywood] depictions of AI. We lose sight of the task at hand. There are some jobs that make sense [to be automated] and some that don’t.”
“It’s about expectation vs reality,” added Harsh Kapadia, executive creative director at VMLY&R. “It’s important to show a robot but more importantly let’s show what it can and cannot do. It comes down to how we apply the notion of AI and how we make it useful. People didn't know what to do with Twitter when it launched, or QR codes for that matter.”
For Radclyffe, we are still a way off from an I Robot scenario. “AI doesn't live up to 90% of the promise. It's only effective in the situations where you have a lot of data and you can only train it to do specific things.”
For example, self-driving cars may well be able to navigate a route and are clearly less likely to make human errors but can’t understand a road sign if it’s been tampered with by graffiti. Furthermore – we are more likely to be horrified by an accident caused by a machine rather than a human – exactly because the more relatable ‘human’ element of error isn’t applicable.
The Democratization of AI
Today, AI is in the hands of a very elite few. Despite Hollywood depictions we’re still in early days with a technology that only a handful of people can build and work on. What will it take to bring it to mass adoption? Is there a challenge in bringing clients on board with AI?
“Take the buzzwords out - they intimidate clients,” stated Kapadia. “Already there is a wall against words like AI and bots. We need to humanise the use of it.”
“Don't talk about the tech, don't talk about the platform. Instead ask what ‘is the problem we are trying to solve?’” added Miller. “Is it going to help you unlock a strategy or help you get the creative solution to where it needs to be? If you don't ask those questions, AI will be like the emperor’s new clothes.”
“AI will only be as good as you are today,” said Kapadia. “We need to set the benchmark. Let's not forget the artificial part of AI. Before we wrote only TV ads or print ads. Now we need to think about being creative or curious in a different way.”
A question of ethics
A member of the audience left us with a poignant question - what is the purpose left for humanity if everything gets automated?
Acording to Radclyffe “there is always debate around models for what the world should be. Philosophers have fought over what the world ought to be for 100s of years. So do economists and scientists today. Inevitably over time - everything that is human could be automated. The question is – should we set limits? What are the human touch things that need to be protected?”
Miller concluded by reframing the question. “Ten years ago most of our jobs didn't exist. Change is coming. As a species we've gone through change before. We need to get free from the protectionist and defensive mindset, have conversations about ethics, ensure we have conversations with clients about ensuring there is humanity in their business. We need to open up the conversation.”