The Agenda with Steve Paikin (Audio) - How Ontario's AI Sector Performs on Global Stage

Episode Date: April 24, 2024

A look at the policies, infrastructure, and people that have made Ontario a top global destination for artificial intelligence research, investment, and talent.See omnystudio.com/listener for privacy information.

From epic camping trips to scenic local hikes, spending time outdoors is a great way to create lasting memories to share with friends and family. This summer, TVO is celebrating the natural wonders that inspire unforgettable adventures with great documentaries, articles, and learning resources about beloved parks in Ontario and beyond. Visit tvo.me slash Ontario summer stories for all this and more. And be sure to tell us your stories for a chance to win great prizes. Help TVO create a better world through the power of learning. Visit TVO.org and make a tax-deductible donation today. Canada, and Ontario in particular, are global players in artificial intelligence.

The province attracts more than twice the AI research funding than other provinces. And AI programs at Ontario universities are world-renowned. But how did we become one of the premier places in the world for AI research, development and investment? And how does it maintain or even improve its position? We discuss this tonight with, on the line from San Francisco, California, Chris Walker. He's CEO of the Toronto-based Untether AI. Chris spent 30 years with chipmaker Intel and commutes between Toronto and San Francisco Bay Area, where he lives. And with us here in the studio, Tom Kaur. He's the CEO of AI Partnerships Corporation.

That's an organization that tries to make AI more accessible to small business. He's also the former CEO of the Accelerator Center Park at the University of Waterloo. Ryan Pereira is CEO of Captain AI, a startup that's automating food delivery dispatch. He's in the process of moving his company to Ontario from London, England, so he can be closer to his delivery dispatch. He's in the process of moving his company to Ontario from London, England, so he can be closer to his biggest client. Here's a hint. 9, 6, 7, 11, 11. Yes, pizza, pizza. And Sanjana Basu, she's an investor at Radical Ventures, an AI-focused venture capital firm in Toronto. She moved to Ontario in 2019 after working for the venture capital arm at Tata Group. What a good group of people we've got here tonight for this. Chris,

thanks for being there on the left coast for us as well. Let's go through this. We know that we've heard a lot lately about how, particularly in the federal budget not that long ago, of federal and provincial initiatives that have helped pave the way for what exists in Ontario today in terms of AI. Tom, get us started here. What kind of initiatives have you seen over the years? Sure. Well, I think the more recent initiatives starting in 2021 with the Pan-Canadian AI Strategy, which is a federal program, saw the establishment of three regional AI centers focused on AI research and working with industry to get AI into the hands of industry for their use. So there's Amy, which is based in Alberta, Mila, which is based in Montreal, and the Vector

Institute, which is based right here in downtown Toronto. So in part, that fed a lot of activity going on in the AI space. And for example, the Vector Institute supports about 700 professors and researchers in the province of Ontario alone in terms of giving them programs that they can offer to their students. So that's a big part of it. And coming out of the post-secondary institutions every year are about 70,000 STEM graduates. So regardless if they go into AI or some other area in technology, you've got a you've got a huge number of people that are in the tech sector. And right now in Ontario, there's 85,000 people employed in the AI sector. 20,000 of those just joined the sector in 2023.

Okay, let's do another example here out in San Francisco. Chris, first of all, tell us a bit about Untether. What is it? Yeah, so Untether AI, we specialize in the design and production of AI hardware solutions that helps make AI run faster and most importantly, more energy efficiently in everything from vision systems to the latest generative AI models. We were founded and headquartered in Toronto. And what tax credits did you avail yourself of in order to help spur on that business? what tax credits did you avail yourself of in order to help spur on that business? Initial credits came from the former, the federal SHRED, Scientific Research and Experimental Program. That's pretty small in terms of the credits. Mostly it goes towards employment.

Silicon's a very capital intensive business. A lot goes into the design and development of chips. Every time you create a chip, it's tens of millions of dollars just in the early production and startup. And so for that, we are venture backed and that's the majority of how we've gone forward, started and gone forward. And one more follow up, Chris, how crucial to your development do you think those tax credits were? we've gone forward, started and gone forward. And one more follow-up, Chris. How crucial to your development do you think those tax credits were? I think it was very important as a seed for sure. Obviously, in the hardware business and in the ecosystem around it, it would be advantageous to develop more, not just in the software ecosystem,

but also the hardware ecosystem. And probably more importantly, as companies like ours develop and deliver our solutions, to have a lot of incentives for the use of AI, for the consumption of AI in areas anywhere from agriculture technology to mining to banking, to really encourage Canadian companies to be at the forefront, not just of developing AI solutions, but actually of utilizing them. Sanjana, tell us about Radical Ventures. What do you do? Yeah, Radical Ventures, we are a global AI fund based in Toronto, Canada. The fund came out of the deep learning ecosystem here in Canada. I mean, as most people know, Canada is the birthplace

of modern AI. Just a few blocks away in the University of Toronto, Jeffrey Hinton and his students released a breakthrough paper on neural nets for the ImageNet Challenge a little over a decade ago. And that set the foundation for the AI revolution that we're part of. And so the fund came out of the ecosystem. It was founded by visionary leaders, in my opinion, who were architects of the AI ecosystem here in Toronto that have set the stage for everything that's happening actually globally today. And our goal is to invest in top AI talent coming out of this ecosystem, coming out of Canada,

as well as globally, to who are solving some of humanity's biggest challenges leveraging AI. And the purpose is to help talent with capital to scale AI innovations both here as well as globally. Okay. My follow-up question is the typical Canadian inferiority complex question, which is you're from India originally, you've worked and lived in the United States, so what are you doing here? Yeah, well, I mean, as you know, I grew up, I was born and raised in Mumbai in India with a deep kind of curiosity and quest for knowledge and new experiences that brought me to the U.S. I went back to India for my MBA, always interested at the intersection of tech and finance. It was actually during my tenure at

the data groups venture arm that I got exposed to the transformative potential of AI, got captivated by the technology, learned that actually Canada was a hotbed for fundamental AI research and thought, what better place to be then and to delve deeper into the technology than Canada. I was also actually, to be very honest, attracted to the openness to talent and the diversity here in Toronto specifically.

As you know, more than 50% of Toronto residents are foreign born. And that adds a richness to daily life and a multidimensionality that you see in all aspects. And then finally, obviously, Radical Ventures was the icing on the cake, like I said, founded by visionary leaders who were architects of the ecosystem, which I was very inspired by, and also mostly inspired by the ambitious vision to be the best AI fund globally. Gotcha. Ryan, let's talk about Captain AI. Is that you? Are you Captain AI? Gotcha. Ryan, let's talk about Captain AI. Is that you? Are you Captain AI? Well, a part of it, for sure. So Captain AI, we're a software company, an early stage startup, that effectively builds software for restaurants to help them manage their drivers.

So if you think of restaurants doing food delivery, we build a software that lets them track their drivers, that lets you as a customer track your driver on its way to you. So you can see in real time the ETA, where they're going. The AI part of our name is that we can actually help the restaurant organize which drivers take which delivery. We do an AI-powered auto-routing and dispatching. So you can take a hands-off approach while orders are being sent to your drivers in the restaurant, the perfect routes are being mapped for them, and they're going out into the world doing these deliveries. And yeah, we've been growing for a while. We've started off in London, England, as actually doing a delivery marketplace ourself, very similar to Skip the Dishes for the viewers here in Canada. We were one of the early players in the UK starting something like that. And then we realized our technology could be so valuable to

other restaurants. And so we sold that business and really took Captain as a software as a service that restaurants could use to help them do better deliveries. Well, same inferiority complex question for you as well. Okay. Starting in Britain, could have gone to Silicon Valley, but you're coming here. How come? Well, there's a couple of reasons. There's a personal reason for myself, which is I got married five years ago, and I moved to Canada because of that. Did you? Are you married a Canadian? I married a Canadian. Cherchez la femme. It's always, okay, love will get it done. There we go. And from a business perspective, I think we've really been encouraged by the Canadian ecosystem. And I think there have been not only investors, but great business opportunities here.

And so I would give a shout out to Techstars Toronto, for example, an accelerator that had faith in us at an early stage. Sunil Sharma, who's the director there, really believed in our company and was able to introduce us to investors in the Canadian ecosystem. Green Sky Ventures eventually invested in us in a seed round recently. believed in our company and was able to introduce us to investors in the Canadian ecosystem. Green Sky Ventures eventually invested in us in a seed round recently. So Canadian investors have believed in our company and really helped us to come over here, as well as our biggest client. We've seen a great spirit of innovation in one of Canada's biggest brands, Pizza Pizza, who again six years ago met with us and believed in us as an early stage company and what we were building. They were able to see that vision. And we've been working with them in a brilliant partnership for six years.

I only have good things to say about their management team who who, again, really foster that spirit of innovation and that ability to, and that desire to be the cutting-edge technology for what they do. Can you say the same good things about their pizza, though? Well, that's a different question. I like their pizza. Okay. Tom, over to you now. The Blue Gene, that's G-E-N-E, Blue Gene Supercomputer. What is that? Blue Gene Supercomputer was a series of high-performance computers developed by IBM 10 to 15 years ago. And at the time, there was a desire to bring the Blue Gene computers to Canada, specifically to Ontario, to give researchers and industry access to them to do things they couldn't do on the computers that they had. So at the time, I was CEO of Ontario Centres of

Excellence, now Ontario Centre of Innovation. So we did a partnership with IBM and funded by the province, brought the Blue Gene computers to Toronto and made them available to people that needed that kind of horsepower in order to do the kind of research, AI and so on. So how integral would you say it has been to the development of the AI culture here in Ontario? Well, it was part of the base of it. Without those high-performance computers, you couldn't do the kind of things that are being done in AI today. And, of course, as time evolves, those computers got replaced by more high-performance computers. But you've got to start somewhere.

And at that time that we brought them up, and IBM brought them up, again, with funding from the province, they were the fastest computers in the world. I want to go back out to Chris. fastest computers in the world. I want to go back out to Chris. You live in Silicon Valley, which I guess everybody knows as the epicenter of this whole thing that we're talking about. And yet there's apparently something about the Toronto ecosystem that you find appealing that you want to be part of it here too. So tell us what it is. I think a big part of it is, you know, we were founded in Toronto and we talked about the university system that, you know, both our founder came out of and also the bulk of our engineering comes out of. So it really is that hotbed of talent across domains of consultancy, software and hardware.

And for us, it's been a great place to have that seed ai or relocate from areas like california other parts of the states we have a lot of people who are moving back home so to speak to canada and because we're located there we're a center of excellence we've got vendors and support uh infrastructure around us really makes it easy for people to say we're doing both interesting work and they see the opportunities to do it in Toronto. Sanjana, let's get an example from you on medical data. Okay, how can AI leverage medical data that is important for doctors and patients?

Yeah, absolutely. Like I said, at Radical, we want to invest in and support companies that are solving global challenges and humanity's greatest challenges. One area that we think that AI will have a completely transformative impact is healthcare. And the reason for that, as you said, is medical data. Medical data is extremely diverse. There are different modalities of this data. It comes in text, it comes in images, it comes in video. It comes in fax still, right? Yes, it comes in fax still. How is it in 2024, doctors are still faxing prescriptions to the pharmacy anyway? I got off track there. Keep going. We do need AI help here, don't we?

Yeah, CDs too. We have a company that replaces CDs called Pocket Health. Anyways, but because you have diversity of data, different forms, different modalities of data, AI is best suited to work with multimodal data to generate great insights, to improve health outcomes. And our hope and belief is that AI will move healthcare from reactive and generalised to proactive and personalised. Ryan, I want to circle back to one of the questions I asked earlier about tax credits that either do a lot or don't do enough to help you get started.

In your experience, how useful were they? Well, I think we're still in the process of actually moving our company from the UK to Canada and reorganizing the company around that. And I think one of the things that's exciting about that is we've been educated about the huge opportunities. And I think I would say it's got one of the Canada, even though we haven't experienced it yet, has one of the most generous programs in the world for giving R&D rebates for staff and for research. We've taken advantage of that in the UK. Traditionally, every year, it's been a great help there. But I think the benefits in Canada are almost twice as generous.

So I'd say it's really great incentives that have been created to bring staff and to bring talent into Canada for early-stage technology. From your perch, does it look that good, Tom? It does, but I think it's one part of the ecosystem that's needed in order for Ryan's companies and other companies to be successful. So getting the tax credits on research that's done is great, but you also need access to the working capital, which comes from the venture capital community.

You need access to the smart people coming out of our colleges and universities. And one thing that we're missing here, and this is more of a cultural thing than anything compared to the States, is they seem to have a lot more repeat entrepreneurs in the Valley. Whether they succeeded or failed, these people will come along and do it again.

And in Canada, if you fail with a startup, you tend to be looked at kind of negatively. In the US, it's, well, what did you learn from that one? Let's get on to the next one. And we don't seem to have as much of that culture here as they do in the United States. So to answer your question, the shreds and the other tax credits

and the funding coming out of the granting organizations and organizations like Ontario Centre of Innovation are extremely helpful, but it's just one part of the puzzle. Sanjana, let me ask you about a short four-letter word that seems to permeate all of this, and that word is hype. How much of what we're talking about today is the genuine article, and there is a genuine belief that this is going to lead us to a better tomorrow,

and how much of all of this is hype? Yeah, absolutely. So, I mean, I work at Radical and I moved to Toronto because I believe that AI is a transformative technology. It is a massive platform shift. So I believe that AI is here to stay. It isn't like a lot of other tech trends that have come and gone. What is hype is the funding

that's happening right now, right? Like the market is extremely frothy. Last year, funding has gone up to close to $25 billion in AI. Is this public and private? This is private investments in AI. Private investments, $25 billion. Venture capital investments.

So the market is a bit frothy right now, but there are funds like ourselves that can differentiate between what is real AI and what isn't AI and what is real AI and what is just.AI in their website. So I don't think that the technology is hypey. I do think that the market forces around it right now has a lot of hype. Brian, you want to add to that? How much hype, how much the genuine article out there? So I have a strong opinion on this.

And I think we're on a tipping point where I think some of the catalysts in the recent years, such as really OpenAI and their work with ChatGPT, has opened up a speed of progress, which I haven't personally seen in my lifetime, where I'm seeing developments on a weekly basis of huge leaps forwards in technology. I don't know if you've had a chance to see the technology called Sora, for example, which is also created by OpenAI, where you've got video generation. And if you look at where

we were with video generation just a year ago and where we are today, it's almost true-to-life video generation that they've achieved compared to something that wasn't close even a year ago. So I think what I'm personally excited about is the speed that I'm seeing these new AI developments happen. I've not seen that kind of speed before. So I actually believe it's, I think yes, there's hype, but I genuinely believe we're at a tipping point where the technology is at a growing and exponential rate right now. In which case, Chris, let me get you back in here to talk to this. And that is there is a lot of concern that has emerged after the federal budget of last week about what particular tax changes, especially as they relate to capital gains taxes, what that might do to inhibit investment in this sector.

And I guess I want not only your opinion on that, but what other regulations that might be in the pipeline that might concern you in terms of inhibiting what we're trying to do here. I think one thing to just start back on the prior is that when we talk about AI, it is very broad-based. It's not just the development of models or what we might interact as consumers in a chat GPT. AI is going to impact autonomous driving, agriculture technology, financial services, healthcare, as we talked about. And we're just starting. So we're just at that point where models now need to shift from being trained to actually being deployed. So you're going to see this next wave of AI come in in terms of actually making it real and real to life for us. From the regulatory standpoint, I think what's really important is because it is so

diverse in terms of the application of AI, that don't have a one-size-fits-all or trying a broad sweeping regulatory approach, that it needs to be domain specific, right? Let people who are focused on healthcare think about the right approach and application of healthcare AI. Same in transportation. So I think it's from that lens that you need to have an approach in terms of how government helps shape or puts regulatory practices in. And then on investment as an entrepreneur,

as an entrepreneurial company, people just think about the founders, right? Or the execs. The heart and the reason people are entrepreneurial the reason startups work um are people basically trade off you know the security or safety in some cases of you know very big companies to do something new to do something different and to risk take and to build and develop. And, you know, the rewards for that are not only seeing your ideas change the world in the case of what we do in AI hardware. Also, there's the financial equity benefit of that as well. And so from that standpoint, the spirit of entrepreneurship

has been proven over and over again means that, you know, people want to see the bulk of the rewards accrue to them and be able to keep reinvesting themselves. And so I think, you know, capital gains and tax policy needs to incentivize that and incentivize the repeat performance of entrepreneurs. And again, it's not just people, you know, executives or founders, you know, the bulk of employees make a tradeoff when they come to a startup for to be more equity based in their incentive. And so when they do succeed, they should they should be able to make sure that they take that in and then go do the next one. Well, Tom, let me give you the example that we've heard most often since the budget came out, the federal budget, that is. Let me give you the example that we've heard most often since the budget came out, the federal budget, that is. And that is something like somebody who wants to make a big investment in AI and therefore, you know, takes the risk of selling a cottage in order to realize that revenue and put it into AI.

But they're now going to pay two thirds. Well, as of whatever, June or July or something, they're going to pay two thirds capital gains tax on that instead of 50 percent, which is what it was. That's an example, I'm told, of a government regulation that will inhibit the growth of this sector. Do you share that view? I do. And I don't think that the federal government came out with the intention to dampen the innovation ecosystem in Canada. I mean, that just doesn't make any sense at all. I'm guessing probably what happened was that we're currently paying close to a billion dollars a week in interest charges on the debt in Canada. And I don't think that the regulators, I don't think that the people who price our debt and the interest rates would be happy with seeing a lot more, another $40 billion worth of expenditures happening through the budget.

So I think what the feds did was look for ways to reduce the gross expenditures by bringing in over five years $20 billion that would have normally gone out in terms of rewards to the investors and to the entrepreneurs, but it's still taken $20 billion over five years out of the system that would have normally, and to the entrepreneurs, but it's still taken $20 billion over five years

out of the system that would have normally, according to their numbers, would have gone to the investors that are taking the big risk. And so there's no question, if you tax something, if you tax gasoline, you do it to be punitive and you bring in a carbon tax to do that, so people buy less gasoline. If you tax the rewards that are available to investors and to the entrepreneurs in building companies, it's going to be a dampener on that as well. So there's no doubt it'll have a dampening effect on entrepreneurship and capital raising, and the risks that our good friends in the venture capital community are willing to make because they're going to get less of a return.

So they'll either invest less or they'll invest it at a lower valuation because they still want that return coming back. So it's harmful. Let me ask our director, Sheldon Osmond, if he would, top of page three, bring up this chart. And those of you here in the studio can see it on the monitors here. This is showing AI investment in Canada in fiscal year 2022. And here it comes. And for those listening on podcast, I'll just read out some of these numbers here. More than a billion dollars invested in Ontario, 600 million in Quebec, roughly 250 million in each of BC and Alberta, and about 360 million for the rest of the country. So add it all up, we're at about two and a half billion dollars

across the country. And yes, Ontario is well ahead of everybody else, but as the biggest province, you would expect that. Sanjana, I guess I want to know from you, are those numbers kind of where you think they ought to be? I mean, ideally, it should be much higher than that. We would like to see much more investment in the ecosystem. And that is why we have, again, like Tom referred to, there's investments that are made in talent through the Pan-Canadian AI strategy which was effectively

a talent strategy for companies like Cohere, Wabi, CentML, Untether, etc. to come out of the ecosystem. There have been investments made in generous tax credits that help companies be more capital efficient and move the money into other facets of the business, like go-to-market. Then you have venture capital investment

that is required to put more money into the ecosystem and into founders such that they can scale. All of these combined are required to increase those numbers. And I actually think we were talking about entrepreneurs in the ecosystem and the American entrepreneurs, how they're different from the Canadian entrepreneurs. I think we all need to invest more and more into this ecosystem because the more you invest in the ecosystem, the more founders come out. If they're successful, they exit. They reinvest their experiences, their capital back into the ecosystem. And that's also the story of the founders of Radical who launched a company, sold the company, reinvested that. And that's our hope with the companies that we invest in

that are from the ecosystem. They grow, they scale, and there's a flywheel because they reinvest in the ecosystem, whether that's their own talent experiences as well as their financial capital. Well, that does prompt a question for you, Ryan, which is, and we did a program about this actually a week or two ago, we don't seem to have the same appetite in this country for risk-taking and upfront investment in AI, venture capital, et cetera, obviously compared to the United States, but even compared to some of the other G7 countries where we should be more on a par, the thinking goes. Why is that? Well, I understand that's a common perception. I don't think I can comment on the industry as a whole because I don't have that experience,

but I could share from my perspective, I've had slightly different experience. I have experienced some of the Canadian ecosystem at least taking risks on us and believing in us as an early stage company. But I guess what's interesting about that is, for example, the example I would give would be Textiles Toronto, which is an accelerator. The director is a guy called Sunil Sharma, and they took a risk, invested in us, and introduced us to the rest of the ecosystem. However, they are originally an American company, so they may go. So it may be that Texas, Toronto embodied the spirit of the

valley where they take risks on early stage companies. But it was through that that we were introduced to other Canadian venture capitalists who did eventually take a risk on us as an early stage startup. All right, Chris, I'll get you to comment on the same thing. You know, you've seen this both from an American point of view and from a Canadian point of view. Why do we seem to be much less interested in taking those risks? I think a big part is when you look at the size of the investments, it definitely takes a lot to compete at the highest levels here. Take chip development as an example.

One chip might take three plus years to develop in excess of $100 million just to bring one product designed to market. And so while the numbers on the prior chart may seem big, they're actually not in the context of actually what it takes to develop and compete at a global scale in AI. And I think one of the big elements of risk-taking is also the community of risk-taking. So we've gotten great backing from folks like Radical Ventures, Canadian Pension Fund, to have the foresight to have those long lead investments. But if you look at tech hubs, Silicon Valley, Singapore, Taiwan, you have that entrepreneurial spirit that feeds off each other. There's movement of people between companies. They support each other.

And one of the biggest differentiators also isn't just the startup mentality. It's also the consumption. It's your banking sector, your healthcare sector, your agricultural sector at the forefront of using and deploying the solutions that are being developed in your AI ecosystem. That's also that flywheel effect that starts kickstarting and feeding off each other because then you want people going back and forth between big enterprise small enterprise to seed innovation ideas and experiences back and forth and so it's not just the entrepreneurial startup risk-taking has to be risk-taking in your key in your key sectors that can benefit from new technologies like AI you need to be at

the forefront of adoption to make it happen. Tom, let me get you to speak to that as well. Insofar as I remember when the budget came out and they said the feds, they were going to put $2.4 billion into AI. And the initial impression from a lot of folks was, oh, that's a lot of money. That's quite a commitment. And then we did a show on it. And people said on the show, yeah, that number needs to be 10 times as big. Are we running the risk of being left behind here? are and if you look at one example is a a good success story right here in toronto is cohere who we're close to and cohere the round that they're doing now reportedly is for 500 million dollars at a5 billion valuation. So that's one company doing one round of $500 million.

And that company's in competition with OpenAI and other large language model companies. But that's the scale of the money that you need to compete in that market. Now that's different than some vendor of software in manufacturing that's doing a few million dollars a year and they want to introduce AI into the product to do quality control on an assembly line which can be done relatively cheaply but if you want to play with the big boys and girls like open AI and so on you

need billions going into the system and I would argue that Cohere will probably need and see billions going into their company in order to compete with the big American large language model companies. Fascinating stuff. And I'm grateful to all of you for joining us here in the studio and Chris Walker out in San Francisco as well. It was great to have you all on TVO tonight. Many thanks. Pleasure. Thank you. The Agenda with Steve Paikin is made possible through generous philanthropic contributions from viewers like you. Thank you for supporting TVO's journalism.