Episode 4: How to jumpstart intellectual property protection in Canada’s artificial intelligence ecosystem

Lisa Desjardins (LD):  You're listening to Canadian IP voices, a podcast where we talk intellectual property with a range of professionals and stakeholders across Canada and abroad. Whether you are an entrepreneur, artist, inventor or just curious, you will learn about some of the real problems and get real solutions for how trademarks, patents, copyrights and industrial designs and trade secrets work in real life. I'm Lisa Desjardins and I'm your host. 

The views and opinions expressed in this podcast are those of the individual podcasters and do not necessarily reflect the official policy or position of the Canadian Intellectual Property Office. 

Canada has five federal technology superclusters. These are non-for-profit entities that are investing federal funds into collaborative technology projects in Canada matched dollar for dollar with co-investment from the companies participating in those projects. One supercluster is Scale-AI based in Montreal, which funds projects across Canada that apply artificial intelligence to solve Canadian supply chain problems. Artificial Intelligence, or A.I. is referring to computer systems that have been trained to make predictions based on data. For scale AI, maybe that's helping to predict truck fleet maintenance to lessen environmental impact, or deciding the best way to unload a container ship to provide the fastest access to urgent Covid-19 supplies. Sounds like cool stuff, but both AI and scaling technology ecosystems come with their own challenges. According to today's guest, Canada's challenge in building an A.I. ecosystem goes beyond developing the cutting edge A.I and includes helping ensure that Canadian innovators are taking the steps they need to leverage their innovations. My guest today is Todd Bailey, Chief IP officer at Scale-AI. Todd is a lawyer, a patent agent, and IP strategist with 25 years of experience of helping start-ups SMEs and multinationals protect and commercialize their IP. And, one of the problems, he says, we need to solve is how to empower Canadian innovators to recognize and prioritize their IP and really put it to work for their business. Todd, it's a real pleasure to have you in our podcast, welcome!

Todd Bailey (TB): Thank you, it's great to be here.

LD: Could you tell me a little bit about yourself and the kind of work that you do at Scale-AI?

TB: Sure! First… you know, I do want to say thank you Lisa for having me here and also thank you to CIPO for launching this series I think it's really an opportunity to give people like myself to share their point of view, so that's great. About me; I guess the thing that I'd say is I love working with people to help protect their ideas. I mean that's the one great thing about this space: you get to meet lots of people with bright ideas - bright people - and it's really awesome. And I've been really fortunate during the time that I've spent at law firms and companies to work really closely with a lot of great engineers and scientists and programmers and business people who are really passionate about what they do. And those experiences have really helped, I would say, shape my perspectives on IP; about what works, and frankly: what needs some attention. You know, where I am now at Scale-AI as you've mentioned, we're co-investing government funds into supply chain A.I. projects as well as into talent development and accelerator programs. And all of that is with the objective of helping to grow Canada's AI's ecosystem. But my focus in particular, is improving what I would call the reflexes for IP protection within the ecosystem. I believe that as a subject IP should be more accessible, more… the information more actionable, and so concretely that means I'm working with SMEs universities and companies to help them see the value of investing their own time and energy into managing their IP. I mean that's essentially what I do.

LD: That's a lot of different roles that you are trying to balance. Here at CIPO we often talk about intellectual property and I think what comes to mind for most people is, when we talk about patents and tangible things: a thing you can touch… But now we're talking about artificial intelligence. What is that?

TB: Well to be honest, the term is an overstatement. I mean we're still really a long way away from thinking machines. A.I. just refers to a series of computer modeling technologies such as machine learning that use data and statistical methods really to make predictions. I mean, it's called machine learning because the programmers don't directly program the model in the traditional sense. Instead they use data to train special algorithms that can figure out the best way to use that data to make future predictions. You know, maybe you studied linear regression in high school math and that's you know fitting a line to a bunch of data points on an xy graph. Well, machine learning is like that except the graph has maybe a thousand more axes than just x and y, so it's complicated. But AI's not science fiction anymore. I mean, your smartphone has tons of AI things like Siri, Alexa, Google voice assistants… that's AI. Image recognition, self-driving cars… there's a lot of technologies today that rely on AI predictions.

LD: Yeah that's true and it's becoming more and more common. So, you're working with still, really, state-of-the-art future uses of A.I. At the same time, you're trying to simplify something that's really important for people and companies that want to succeed… and I've read your blog, "IP without Jargon". Here, you explain IP concepts, debunk some IP myths and teach some IP terminology. I love the work that you do there to try to simplify and really convey the message that it's important to protect your IP. Why did you decide to create this blog?

TB: I'm focused on some disconnects that exist in the Canadian start-up culture that I that I want to help address. There's a well-known problem: that Canadian SMEs are lagging their global peers and you know, maybe two percent of Canadian SME's own patents according to studies. And that's compared to almost 10 percent for, you know, south of the border in the U.S. and… but it's more than a patent issue because the patents are really I think a canary in the coal mine of symptomizing a larger problem. But at the same time, you know, there's a lot of evidence that SME success is linked to protecting IP. And so: Why aren't more start-ups protecting their IP? That's the problem! And for me, I see two main barriers at work: what I call a knowledge gap, and an emotional gap. But the knowledge gap isn't what you're thinking: it's not a lack of information. In fact, IP information is everywhere. What's missing is a connection between abstract ideas and everyday life. Like: how does this apply to me? And does it even apply to me? I mean we're living in a time where information isn't trusted anymore. People are more suspicious about the motives behind the information, and that's where the emotional gap appears because you know changing behaviors is hard. Like how many new year's resolutions don't make it to the end of January? I mean for me, it's probably the first week of January… so, to make a change you know you really have to believe that change is necessary. You really need to appreciate why you need IP before it can really help you: it's not a checking the box kind of thing. The emotion is seeing a benefit and really wanting it for yourself. So the blog is really just me reaching out to try to help tackle these issues.

LD: Yeah I'd like to read a section of the blog where you debunk myths. It says: "IP myth. IP is old-fashioned. Today, success is all about defending moats." And the IP facts says: Buzzwords change, but good business practices don't. Protecting your digital IP is more relevant than ever before. Your ideas, information and data are only valuable as long as you are able to protect them from competitors." You're talking a lot about how to protect. And the gaps, that you mentioned, that kind of prevent people from thinking about protection and understanding the importance of it. So when it comes to protecting IP and the projects that Scale-AI is funding: what are the main lessons or concerns that you discuss with the engineers and programmers and so on in the supercluster?

TB: Listen, there's a lot of confusing information and even conflicting opinions out there. So, you know, I respond by: I try to be simple, and clear: If you operate a business that's innovating, but you're not protecting your IP, your days may already be numbered. No one's copying failing businesses. But if you're successful, if you're winning, you're going to attract your competitors and your copycats. And they might even be better funded than you, or have fresher legs to outrun you. So I like to ask people, you know: What are you going to do when another business shows up with a similar name, doing similar things? What are you going to do when a large incumbent sues you for patent infringement? You know, the message is that rainy days are going to come, and if you don't have protected IP and an IP plan, you just have fewer options for survival it's really that simple… Eventually your time is going to run out. Because you know history is full of stories about these kinds of preventable failures. IP is actually quite unforgiving; you really only get one chance to protect it, and so, I like to say, you know: IP delayed is usually IP denied.

LD: I agree and that's something that I know that our people see as well: you really only get one chance…

TB: Yeah…

LD: Speaking from your experience: give us some of the key IP takeaways that you can share with creators and inventors.

TB: You know I guess there's three main messages I try to get across… again, trying to keep things simple. The first is that IP is a business tool, it's not a legal tool and it's really not that complicated. Especially if you strip back all the jargon the concepts are actually pretty simple and logical. Yes it's true, IP does involve lawyers, just like when you buy a house. But there's no one who lets their lawyer decide which house to buy, or how much to spend, right? And so you need to see your IP in the same sort of way: you need to own the decision making process to really benefit from your IP. The second message that I try to get across is about, you know, protecting IP doesn't have to be expensive: there's ways to prioritize your spend and to control costs. But these aren't even costs; these are investments that you're making in the future success of your business, and investments to improve your odds for success even. And so along those same lines: my third message is really that you do need an IP plan. Some people call that an IP strategy or an IP roadmap, but I like to call it an IP plan to sort of highlight that actions and planning are involved. And actually, you know, start-ups write all kinds of plans: they write business plans, and social media plans, recruiting plans, and so on… The reason's simple, really: because the plan helps you think out the approach ahead of time so you can be proactive. Nobody writes a recruiting plan after they hire and… you know, so, IP needs that sort of same mindset. And listen, today innovative businesses without an IP plan are really just showing their ignorance to potential investors. But you know, it's not so simple because coming back to the knowledge gap I mentioned a moment ago: there's a lot of white space between knowing that you need an IP plan, and having one that suits your business. That is like moving from the abstract to the concrete: that's the real challenge.

LD: Yeah let's talk about that white space for a moment. Because you're saying that IP isn't complicated, it doesn't have to be expensive, but it does require planning. So, in terms of jump-starting the IP protection in the Canadian AI ecosystem: what do you think is that white space: the stuff we don't see?

TB: You know, that's a great question. You know, for sure we have to keep working at getting the big messages out, to close the gaps. But beneath the surface you know.. I would say that that one thing that's not always seen is: there's a strange paradox that exists between IP and AI that that creates some headwinds. As a scientific technique, A.I. has been around for a long time, almost 70 years if you can believe that. But the science has had to wait a long time for technology to catch up for faster computers cloud computing and so on. So you know, A.I. is kind of old and actually patenting the first patents on A.I. methods are actually also pretty old; they were filed decades and decades ago… So, yet within the ecosystem there's a persistent misconception that A.I. can't be patented, but it definitely can be patented! There's a patent boom going on right now and according to WIPO, the World Intellectual Property Organization, globally, there's about 300 000 A.I.-related inventions, either patented or in the patent process and that's huge… So why does this notion that A.I. can't be patented still persist? It's a tough one, I'll be honest with you. But that paradox really highlights one of the biggest IP challenges facing us, and we really have our work cut out for us in trying to address that. But at the same time I'm also going to say that, you know, there's more people than ever focused on this problem. So I'm really hopeful for the future; I really think we're heading in the right direction.

LD: Todd, it's been a pleasure and I'm really happy to have you join me on our journey to try to educate people on how IP works and why it is important. Thank you so much for taking the time to talk to me!

TB: It's been my pleasure, thank you very much.

LD: Thank you!

You've listened to Canadian IP voices where we explore intellectual property. In this episode you met with Todd Bailey, who is a lawyer, patent agent, and IP strategist. Todd explained some of the myths and challenges he is facing in his role as Chief IP officer at Canada's supercluster Scale-AI. Todd is on a mission to jumpstart IP protection in Canada's AI ecosystem. To read his blog "IP without Jargon", go to scaleai.ca/blog/