For more than two decades, artificial intelligence and machine learning have been playthings for Amazon.com Inc. The digital retailer, after all, has long been in the business of predictive recommendations for consumers. With its cloud-computing subsidiary, Amazon Web Services, Inc. (AWS), the company is now in the business of selling AI and machine-learning capabilities to other businesses.
Glenn Gore is AWS's "chief architect" – a title that, in his words, means spending a lot of time in airports. It's the London-based evangelist's job to sell the world on AWS's capabilities, with AI and machine learning among its biggest frontiers. Last quarter, the fast-growing division's revenues increased 42 per cent to $4.58-billion (U.S.), contributing to Amazon's growing bottom line. The cloud service is rapidly expanding in Canada, as well: its operations include a Toronto office and a data facility in Montreal, and the company has 99 job openings across Canada.
In a wide-ranging December conversation in Toronto, Mr. Gore discussed the potential AI and machine learning could bring to Canada and the broader business community.
What is it that you do?
It's really about channelling the collection of knowledge back into this single set of best practices, and then sharing that with customers. That could be doing industry events and raising awareness or deep dives with customers about how they're going to change the organizational culture of their IT department. There's no real restriction on the role. It's every industry, every customer, every country and that variety is just incredible in terms of connecting the dots together. We sit down, just talk about how all this is working to make sure the customers aren't going to be negatively affected going forward.
How can more companies adopt AI and machine learning?
The real challenge is, how do you take an idea and turn it into something real? It can be really overwhelming. How do you keep AI talent? What happens if they leave? This space is innovating so quickly, what if you back the wrong technology? Are you going to be exposed? But if you're looking about how to apply machine learning to your business, you don't need to know everything. You can use some of the AWS services we've talked about, and you can build up pretty complex use cases without having to understand how the machine learning itself is working. You're just interested in the application of the tool to your business. Get started, experiment, come up with that business idea, then innovate around it. Have a hack-a-thon, bring interesting people in – you'll be amazed what you can do on your own using some of these out-of-the-box solutions.
Tell me about your relationship with the Canadian market
One of the reasons we invested in the Montreal region is around the locality of data: every country around the world wants to keep data within their borders as much as they can. But it's one thing to put the technology within Canada, and it's another thing to invest in the solutions architects, professional services, training certification teams, the partner organizations. They're out there working on a daily basis with customers, partners, education bodies, government departments, helping answer questions. How do they leverage these technologies and bring them into their organizations? How do they retrain their IT workers? How do they bring this culture of experimentation back in here?
So you want to enable more research and development?
What the cloud has really brought back to the IT industry is it's okay to try things again. You don't get punished for something not working. Whereas in the past, if you went and had to spend a million dollars on physical IT equipment and a data-centre contract, if it didn't work, you'd probably be out of a job. Whereas in the cloud, you know what – spend a week, give it a go. Often it'll cost you nothing because it's in the free tier of AWS, and if it has cost you something, it's normally less than a cup of coffee.
The biggest shift that's happening in some industries – they had outsourced IT. They bought a lot of off-the-shelf components. And while that worked for a period of time in terms of reducing costs, it ultimately came with a longer-term cost: it removed a lot of their ability to innovate or differentiate. And what you're seeing now is a lot of enterprises coming back to in-sourcing: how do we hire developers? How do you bring a competitive advantage through our use of technology? We want to make it as simple as possible for as many people to start experimenting with machine learning.
What Canadian AI innovations do you see getting bigger?
Some local Canadian customers are at the start of what's going to be a pretty interesting journey. [Kitchener, Ont.-based Miovision Technologies Inc.] is a great story, where they're using the Internet of Things across thousands of devices looking at traffic management. If you look at what they've been able to put out to date, it's revolutionary. But if you think about all the things they could do over the next five years using this technology, they're going to be incredibly successful. Ambyint helps manage oil and gas wells – they're looking at the Internet of Things, using connected sensors and artificial intelligence for prediction. When is something going to fail, and can we fix it beforehand? That has a huge impact on the environment.
We're in Toronto right now – where do you see Toronto and surrounding region in the global AI perspective?
Toronto has all of the components to successfully be great as an AI hub. It's got talent, it's got great education systems, great quality of life, it's a good city, it's connected, it's got businesses around it. It's got a good startup ecosystem being built around it. There are all the ingredients you need. So you have all of those components. You're starting to get the organizations – like the Vector Institute and Creative Destruction Lab – now it's just about doubling down on that. I think part of it is marketing. Don't be shy of shouting from the rooftops the great things that happen in Toronto. It's a little too shy.
One consequence of slow adoption can be brain drain. How can a growing market keep talent?
I was born in Australia; Australia and Canada are very similar. Just flip it around the other way and swap your bears and wolves for snakes and spiders, and you pretty much have Australia. We kind of learned to embrace the brain drain. Talent exiting isn't actually a bad thing, because they're learning a whole set of skills and experiences. We see this pattern occurring: You'll grow up in a country, go work somewhere else for a couple decades, but often you'll come back home. And when you come back home, you're getting these incredibly talented individuals bringing skills back in, and looking to mentor the next generation.
This interview has been edited and condensed.