In the second part of our interview with Chris Ford, manager of the Smith and Williamson Artificial Intelligence fund, we discuss the potential transformation of urban landscapes, healthcare, regulation and why he doesn’t hold IBM. (In the first part we discussed how AI winners were set to establish an unbridgeable lead).
GIM: How is AI set to change the urban and transportation landscape?
Chris Ford: “If you take transportation and logistics, the revolution that’s happening is transformational.
“In California over 90% of the miles driven have been by automated vehicles also happen to be electric. Once we get to a world where automated vehicles are our reality, it changes the economics of transportation and cities, beyond all recognition.
“So many of the costs associated with traditional ride sharing come from the guy who sits in the front seat. You’ve got to pay him and insure him. If you remove those costs you radically cut the costs per mile, and if you drive down the costs the attractions of giving up your care, become more apparent.
“None of this can be delivered without an AI platform. An Ocado delivery and automated packing facilties can’t work without AI and automated driving can’t work without AI.
“At the moment, automation sits outside personal experience. It involves futuristic concept cars in California – or a dude in a Tesla filming himself on YouTube – little in downturn areas.
“But within the course of the next 12 months, we’ll have automated taxi fleets up and running in Phoenix, Arizona and will have them working in downtown San Francisco.
“The Tokyo Olympic Games will be you will be driven around by AV. It will go from being out there on YouTube and in a kind of Mad Minds to being part of people’s lives.
“We will be quite surprised not necessarily how quickly we adopt them perhaps not over the next 12 to 18 months, but certainly in the next five years.”
GIM: How transformational will AI platforms prove to be in healthcare?
Chris Ford: “Healthcare is front and centre of the change. It is going to transform the way that we are provided with healthcare and by extrapolation, the way we lead our lives.
“We will be provided with more information, a consumerisation of health care. Often in the past, that has been seen as some kind of dumbing down. Yet if you can provide more data to people on a regulator basis, you can keep them healthier, things they can do for themselves.
“AI platforms can also completely change the way medical imaging is delivered. Error rates among radiographers in the NHS are still not inconsiderable up towards 10%. If you deploy artificial intelligence platforms, in many cases, you can significantly reduce that.
“If you use both the radiographer and the AI together you reduce the error rate still further – a great of example of how you are not going to fire the radiographers but ask them to work with the AI.
“You can use AI do very significant image enhancement. What you really want is a really high definition image, but if you’re using radiation as part of that image taking process, you’re limited in the imaging that you can even achieve and the patients who can use that imaging.
“You can’t use high doses of radiation to take really sharp images of infants. But now you can use AI to significantly enhance images, which would be completely unusable by humans otherwise.
“We should care about this, because it gives us better healthcare, we should also care because it reduces the cost of health services.”
GIM: You don’t hold IBM despite its reputation. Why not?
Chris Ford: “It’s really interesting stuff that comes out of Watson and they’ve got a great heritage but I can’t buy Watson now. I can only buy IBM. Most of what IBM does is services. They sell advice, consultations and services to people who want to understand better how to manage transition into the digital world. But there’s a lot of a lot of stuff that relatively speaking is blocking and tackling, just keeping the lights on. From my perspective, it’s more a case of you looking at what they do.
“They spend more time talking about AI than almost anybody else. But we think the lady doth protest too much. There are other relative taciturn companies that are at least as well positioned or arguably better positioned than IBM with regards to AI and much more purely exposed.”
GIM: What about the FAANGs? Does their AI reliance mean they can be classed together?
Chris Ford: “They may share some characteristics, but candidly, not very many. It was a pithy way of describing a bit of the market for a period of time. I’m not convinced necessarily that it actually adds a lot, when looking at them all together.
“What they do share is that they’re very large, and they’re all engaged with AI. It is at least vicariously interesting to observe that this group of extremely large and, by historical standards, rapidly growing mega cap companies, all of them are emblematic of artificial intelligence platforms in one way, shape, or form.
“With Alphabet, you could say that there’s really nothing they do that could be done without the AI. It’s also interesting to what does it look like in the Asian market? The companies we bring in when we look outside of the North American context are Tencent and Alibaba.
“What’s interesting about all of these companies, and they’re all building platforms is the infrastructure, on which AI might sit.
“While there’s not really much in common between Microsoft that sells productivity tools and operating systems to consumers, but largely to enterprises and Facebook, which basically is a social network, they both sit in the upper echelons of the market. They are both stuffed to the gunnells with AI. I fight shy of drawing too many kind of commonalities with respect to the markets they address, the particular corporate structures and also their attitude towards data governance and ethics.”
GIM: How much do you have to focus on regulation?
Chris Ford: “The first thing to say is that regulators are not up to speed and you can absolutely be sure that the expertise and resources inside the big companies is significantly in excess of the resources available to regulators.
“The distinction between regulation of big data, and the regulation of AI is a bit of a moot point. Ultimately, it ends up being the same thing, not because Big Data is the same thing as AI but because you can’t run artificially intelligent platforms without the data that feeds them.
“Big data is not a sufficiency for an artificially intelligent platform, but it is a prerequisite. So by regulating the prerequisite, you end up regulating AI. I think it is appropriate that in the first instance, governments look to focus on data, how it is stored, kept, manipulated, and moved rather than specifically at the kind of the AI angle. It’s easier for a regulating body to define what data is rather than trying to define what AI is.”
GIM: Is this still a world dominated by California?
Chris Ford: “What we’re seeing around the world is a real kind of broadening of the innovation profile. Look at the 1980s, Japanese were leading in their involvement in technology Sony, Sharpe, Toshiba. Go back to the 1970s and that was the case for Europe.
“So there’s nothing in Newtonian physics that dictates that innovation has to be something that happens on the west coast of the United States.
“This isn’t to say that US firms won’t have a role to play, they will. But it is to suggest that if you don’t keep your eyes peeled, you’re destined to be surprised when another company bubbles up from a different part of the world.
“Something which very much helped the US over the course of the last 30 years is having access to this vast homogenously governed market, you can launch into in Texas or Delaware. That is available in China, this huge internal market, which of course the European Union to trying to create too.
“As China gets better at indigenous innovation some of these companies will bubble up to be to a real scale as a the result of their access to this faster, digital market.”
GIM: Is the world dividing into spheres of influence between the US, EU and China?
Chris Ford: “Ultimately, that’s how it’s going to pan out. The geopolitics will determine things broadly speaking, how it has to fit the country. If you are Vietnam, you’re probably gonna deal with the Chinese regulatory framework. It is more interesting for places like Australia, for example and the UK as well, if we’re going to follow GDPR post Brexit or something different.
GIM: Does AI carry the threat of more unemployment?
“It is the rightful place for government and regulators to be thinking about this and we need to understand what the regulatory regime will look like. It will be much more draconian, if governments believe that this technology is going to be significantly disruptive to society, the law and from an employment perspective.
“My view is that it will be transformational from an employment perspective but there’s no reason to think that the Luddite fallacy is any less fallacious today than at any point in the last 250 years.
“It is a challenge for each generation facing each succeeding technology wave, to figure out how you deal with the creative destruction and destruction.
“There is no evidence from the historical record suggests that we should be particularly more worried this time.
That’s not to say that all countries or regions or companies will be untouched, and individually, if you are doing those relatively menial, repetitive tasks in a white collar context, then you can be sure that it will affect your job at some stage.
“The question for governments and regulators is how do we cater to those people who are displaced and disenfranchised from the workplace as a result?
“It’s another means by which countries will differentiate themselves from another, the way which they respond to challenges, but I don’t think we should be fearing that unemployment is the obvious conclusion to the deployment of artificial intelligence platforms.
“When I speak to companies who are deploying robotic process automation tools, I have yet to find a single example of a company that has implemented a move to these platforms and that has done so for simply for cost reasons. Almost all of them do it, because it allows for greater accuracy, and greater timeliness. If you think about doing that working in a regulated context in financial services or as healthcare professionals, if you can avoid regulatory breaches and do the relatively mundane stuff on time, it is a no brainer. And that’s got nothing to do with firing people.”