I spoke to applied futurist Tom Cheesewright (AKA Mr. AI) last week for our latest episode of Call on me (listen above!). He was eagerly navigating his way around London at the time, hence the background noise and general hullabaloo.
If you haven’t got 18 mins to spare right now, here’s a quick rundown of our noisy chat all about the current state of AI and how scared we should be of it. Over to Tom…
SCARY ARTIFICIAL INTELLIGENCE?
“AI is a technology like any other that can be hugely abused in its application. The more reliant we are on the technology, and the less we consider its behavior, we start to see examples of what happens when algorithms run amok. For example, the UK stock market flash crash in 2016 was partly blamed on rogue financial ‘algos’ trading on the back of inaccurate trending news stories. So mistakes can happen even when the algorithms are only slightly wrong.”
LUDDITES: THE SHORT-TERM THREAT
“I think there is a very real threat in the short term, which is the luddite threat. The luddites were never scared of technology. They were scared of technology taking their jobs. So the most immediate threat is absolutely the loss of jobs, and people are very right to be scared of that because frankly it’s going to happen. But the potential dangers are outweighed by the benefits.”
STAGING AN INTERVENTION
“Intervention is always cheaper and more effective than cure. Whether it’s a social situation, a health situation – cancer or asthma, whether it’s domestic violence, obesity or mental health issues, the earlier you can intervene and make a successful intervention, the more successful it is for the patient or person involved and the cheaper it is for the state. So if you can apply AI to those sorts of situations – for example, multiple call-outs by environmental health to a property could be an early indicator of a social care problem- then the outcome could be so much better. The same goes for cancer. You throw historical data about cancer diagnoses at a machine to spot patterns, you may find that there are early indicators that no doctor has ever recognised. Cognitive jobs that might take a thousand researchers a thousand years, machines can do it in a matter of days.”
“Some real-world practical applications of AI could make our lives so much easier in future. Take the property sector. This is a ‘high friction’ industry, involving huge amounts of paperwork. I did a remortgage recently and I found myself turning into the Incredible Hulk on a regular basis, just screaming at the computer. It was such an unnecessary waste of my time that these deliberately appallingly designed processes, designed to maximize the amount of time that the lawyers had to spend on it to maximize their billings. Imagine replacing this process with AI: one on my side, one on their side. My AI holds all of my documentation, my ID, and can validate my identity on my behalf, and then on the other side you’ve got an AI operating on behalf of the lender that puts in requests to my AI, sends notifications to the borrowers phone, validated through thumb print. The entire transaction could be over in 15 seconds.”
OUTSOURCE TOILET ROLL SHOPPING
“A lot of the stuff in our lives gets automated now. No one enjoys shopping for toilet paper or tins of tomatoes. But we all have to do it. So we could outsource this. Discretionary purchases will be outsourced to an AI before too long, you’ll just let the AI get on with those bits of your shopping. This means that you will do the fun stuff, buying the steak and the red wine, the flowers and chocolates. So markerters will in the future have a whole new set of signals to analyze. They will have to understand how the machine is making decisions and what signals drive those machine-made decisions. But you’re also always going to be using creativity to engage with human beings. The nature of marketing is always going to return back to creative. The creative application of technologies remains very human.”
“You have to look at what is happening both at machines, and human beings. Machines are getting better and better at filtering out machine-made content. I don’t see spam anymore; it never enters my inbox. So machines are very good at identifying the stuff that I just don’t want to read. And look at human beings. The nose for authenticity that your average 21-year-old has now versus the gullible-ness of an average 40-year-old- I think there’s a better human and better automated ability to sniff out stuff that isn’t human-made. This gives me confidence that the stuff we’ll talk about, the stuff we’ll remember, in the future will always be human.”
Well said Tom.
You can find out more about Tom’s work below.