Alan Turing never thought of this. The genius who helped to break the wartime Engima and Lorenz codes, and who laid down many of the templates for modern computers, once theorised that you could only decide if a machine had become truly artificially intelligent if you could have a conversation with one that was identical to one you would have with a human. The Turing Test.
I’m not sure Turing ever envisaged part of that test might be the computer trying to sell you a second-hand Ford Focus. Very nice alloy wheels, sir or madam. Full service history, sir or madam. Death to all humans, sir or madam.
But the day when a pile of circuit boards can tease the wallet out of your pocket with blandishments about reasonable finance terms and an inclusive service package might not be as far away as we might think. In fact, according to one AI expert, the machines could be selling us cars (machines selling machines is surely a pass for the Turing Test) as soon as 2025.
“It’s only a matter of time before the technology advances to a point where the likes of ChatGPT can get very close to how a salesperson sells cars offline. We think this could be as soon as 2025 based on the current speed of AI development and adoption,” says Johan Sundstrand.
You might argue that Sundstrand – the chief executive and founder of Swedish video tech pioneer Phyron – would say that, wouldn’t he? The company has been developing its own AI-powered software, and has designed it to be able to identify the best and most relevant selling points of each individual car and combines still images with a data feed, facts and figures about the car, brand imagery and retailer services into what Phyron calls “relevant, highly effective videos”.
“We know the automotive retail market is more competitive now than ever before. In this environment, anything that gives your business a competitive edge and increases efficiency should be welcomed. AI-powered videos have proven to sell cars three to five days faster, and the addition of AI-powered voiceover will only increase impact and efficiency,” says Sundstrand.
Relatively simple AI tech such as this could be the thin end of the unemployment wedge, unless its creation and deployment is handled incredibly delicately
This rather seems to ignore the potential for mass unemployment in the car trade. In Ireland, some 30,000 people are employed either directly or indirectly in the process of selling and servicing cars, and a fully enabled car-sales chatbot could take an awful lot of those jobs out of commission. And, indeed, wouldn’t need to be paid commission.
If that sounds apocalyptic, then the worrying thing is it’s already under way without any AI in sight. The business of selling cars is moving rapidly to an online model, especially as car makers start to switch over to the agency model of selling their wares. In this, dealers no longer buy stock from car makers, but simply act as a fixed-payment intermediary, or agent if you prefer, for the car maker, allowing for more of the sales process to take place online. Tesla already operates this model (in fact, it goes one further by also owning its own outlets, rather than relying on franchised dealers) and the likes of BMW and Mini are due to follow suit, shortly.
Of course, Phyron’s tech can’t quite create an automaton salesperson, clamping your fist in a metallic handshake while passing you a pen to sign on the line. Not just yet, anyway. For now, the Phyron tech is more about creating online content that helps to sell the cars. Its algorithms enable Phyron to create videos for car advertisements that can be used on etailer websites, as well as classifieds, across social media channels and targeted email distribution.
[ Chris Horn: AI is getting smarterOpens in new window ]
Not only do automated videos potentially save retailers huge amounts of time and money on video production, especially as there is no need for a salesperson to video any cars on site or prepare a script of key features, but Phyron claims they ensure premium-quality films of every single car in a firm’s inventory has an engaging video ad attached to it in minutes. Each advert takes about 10 minutes to create and render.
The process works by Phyron’s tech using a data feed from a car dealer, including at least three still images and key product details. The AI then identifies the key properties in the images, including various shot angles, interior elements and exterior highlights, and then strips out the backgrounds and replaces them with a neutral one. The software can then add special effects, shadows and overlay the desired product specifics and any customer offers on the relevant images, including prices and retailer dealer branding.
It’s happening already. Phyron says it has 1,500 clients across Europe for this video-making tech, adding up to 100,000 AI-made adverts for cars. The Swedish company claims that, since January 2021, its customers’ car videos were viewed more than 250 million times.
A report by Goldman Sachs earlier this year suggested that generative AI could replace the equivalent of 300 million jobs globally
Relatively simple AI tech such as this could be the thin end of the unemployment wedge, unless its creation and deployment is handled incredibly delicately. A report by Goldman Sachs earlier this year suggested that generative AI – that’s AI that is truly indistinguishable from a human, blowing the Turing Test out of the water – could replace the equivalent of 300 million jobs globally.
Speaking to the BBC, Carl Benedikt Frey, future of-work director at the Oxford Martin School, University of Oxford, said: “The only thing I am sure of is that there is no way of knowing how many jobs will be replaced by generative AI. Consider the introduction of GPS technology and platforms like Uber. Suddenly, knowing all the streets in London had much less value – and so incumbent drivers experienced large wage cuts in response, of around 10 per cent, according to our research.
“The result was lower wages, not fewer drivers. Over the next few years, generative AI is likely to have similar effects on a broader set of creative tasks.”
Phyron’s tech is not generative AI, but it’s a step on the path to that kind of technology. We’ve known for decades that robots can replace us at our jobs, if we’re not careful, but who knew that car dealers would be the first to suffer?
