Within the past two years Montreal has become a global leader and talent hub for artificial intelligence (AI). But Sylvio Drouin started a company in Montreal focused on AI and machine learning 17 years ago. For someone who invented a computer language on a typewriter before he could get his hands on a computer, it shouldn’t be surprising that Sylvio has never let the fact that he’s usually leaps and bounds ahead of the technological curve slow him down.
Currently the Vice President of the Research Labs arm of Unity Technologies in Silicon Valley, this former Montrealer has been around the world and back again–whether it was working as architect to an entire tv satellite station in the middle of the Malaysian jungle or founding an AI company in Montreal before machine learning was widely known.
“I have an interesting background in the sense that I taught myself programming when I was 10 years old,” Drouin told MTLinTECH. “I didn’t have a computer, I had a typewriter, and I ordered a book on Pascal. I learned to read English and to code. I invented a computer language on the typewriter because I didn’t have a computer. Two years later my parents bought me a computer and I started to code furiously, ferociously.”
He attempted a semester of university majoring in chemistry, but dropped out soon after. He was modifying operating systems full time for computer companies by the time he was 16.
“From there, I started working on graphical user interface. Then I joined Eicon Technology as lead architect for telecommunications systems. I worked on large telecommunication infrastructure around the world for Lufthansa, Air France, Chase Manhattan bank, NATO, all these large infrastructures.”
During his nearly seven years at Eicon, Drouin wrote the initial whitepaper that led to the development of the Aviva Terminal emulator that sold several million copies worldwide. He was was also the principal architect and lead developer of one of the first virtual reality based network visualization tools as part of an R&D effort in the company Ireland labs. Eicon was later acquired by Dialogic for $400 million.
“From there I was living the life of a rockstar in the sense that I was partying and hacking and being paid tons of money for it and traveling around the world. I had this crazy lifestyle. I ended up in Malaysia, where I became CTO of a satellite tv consortium founded by Japanese, American, and Australian investment, and I was the architect for an entire tv satellite station in the middle of the jungle in Malaysia. It was a really crazy experience because there’s a lot of corruption there–it was in the initial phase of the prime minister trying to make Malaysia a hub of technology throughout Asia. We realized and implemented part of the architecture, but it was really a tough environment.”
In January 2001 Sylvio moved back to Montreal and founded MAZ Laboratory inc. with $5 million investment from Japanese investors. MAZ was founded to work on machine learning and genetic algorithms: a full 17 years ahead of the current AI trend.
“I was doing visualization and visual programming tools mixed with algorithms to expand the experience over time. It was really crazy because I was far too ahead of my time and people had no idea what I was talking about. So I sold back all the technology and the patents to investors, and I think they sold it to Fujitsu in the end.”
Drouin spent the next year in New York completely immersed in the art scene and creating art. After which he moved to San Francisco and became the CTO of the one of the first mobile photo-sharing companies, Tiny Pictures Inc., which was later acquired by Shutterfly. Then the moved on to CTO at Extranormal, also in the Valley. Extranormal was one of the first 3D storytelling tools on the web, and was also based partly in Montreal.
Before he left Extranormal at the end of 2011, Drouin had started working as an advisor to Unity Technologies, one of the most successful game engines in the world. He’s been with Unity full-time since for the past three years leading all the advanced research. The company has worked on VR and storytelling with partners like Google, Facebook, Snapchat, Apple, and Samsung.
“I do strategy for the company, I hire all of our rockstars, I work on some of the most fundamental core technology in the company. And we’re pre-IPO and it’s a big crazy adventure.”
As someone who has spent a good amount of time both within and outside of the Montreal tech ecosystem, he has seen its global reputation grow over time.
“I would say Montreal is, after Silicon Valley, the most amazing tech hub in the world in terms of talent. If you look at the game development in Montreal, the core talent in deep learning and machine learning, and the core talent in telecommunications–it makes Montreal one of the most appealing places in the world to grow a company and grow a team of talented developers. It wasn’t always like that, it became like that. Now more so than ever.”
“Montreal was already good in graphics 17 years ago, and since Ubisoft came to Montreal, they’ve created this crazy gaming industry with lots of talent. And then machine learning, between 5 and 10 years ago universities started to invest more in machine learning. And that created a lot of talent, and then Google spotted that, and now every company wants to be in Montreal because of it, for the machine learning knowledge and talent pool.”
Drouin describes Montreal as a name known to a select, in-the-know few in the Valley. But the people who do know about it are key players. And in New York it’s come to be common knowledge that Montreal is the place to build a tech business. He’s even started to see startups and companies moving from New York to Montreal, because of the talent and the lower cost of living and the knowledge.
For now, Sylvio is heading a lab of about 30 people spread between Silicon Valley, Montreal, and Grenoble working to create and influence the future of game authoring, rendering, VR, AR, and storytelling.
“We’re working on skin, eyes, hair, and facial capture and facial animation. So basically we take AAA facial capture and bring it to smaller indie studios and developers. We are also using machine learning to make eyes more human-like. We’re working on a concept for smart asset, because when you assemble a game now, your assets are really just geometry and textures and shaders. Every time you make your asset smarter in the pipeline, in our tools you peel one layer of complexity back on the pipeline. It means you decrease the complexities of making your game every time you make your assets smarter, and how you make your assets smarter is that they know how to connect with each other. A wheel knows it connects to a car, and the car knows it’s transporting things from point A to point B. The car knows it’s a road. The game developer will not have to add that behavior to these assets. The assets will know how to behave amongst themselves and how to behave over time.”
All of which works to decrease the complexities in making games. The lab is also working on mixed-reality, and the future of how content will be created and consumed with mixed-reality.
“We are looking at authoring in VR, from a consumer angle and from a developer angle. How do you make authoring experiences in the VR environment itself, not outside of it? We are walking on the future of AI, the future of applying VR to game character, and to planning.”
“We’re borrowing a lot from what NASA did with the Mars rover, and self-driving cars, and robotics around the world, and we’re taking those high-end technologies and trying to bring them into game authoring and game development.”
“And soon we are going to be doing research on storyboarding and story mapping, so we can help people that are downloading Unity to create stories or film on how to get their ideas from initial seed to an actual product and an actual piece of content that you can publish.”
Outside of the research lab, Sylvio also works at Unity across various strategic initiatives, including acquisitions, growing the company, and hiring across the entire company.
“A lot of our time as execs of this company is spent hiring top talent. Silicon Valley is such a competitive place you consistently have to be on it all the time. And I’m traveling around the world because we have offices in 23 countries.”
So what are Sylvio’s predictions for the future of VR considering he was 20 years ahead of the AI curve?
“With VR there’s a lot of work to be done. Because VR is a multi-dimensional storytelling platform that nobody fully understands yet.”
“We are about as far in our understanding of VR as people were in film 100 years ago. It took people a while to develop a film language, so that when people go to movie theatres to watch movies. With VR we’re at the same place.”
“There’s no compelling content yet. And that is why the VR numbers are just not there. Now augmented reality (AR), with ARkit and ARcore (Apple and Google’s respective AR frameworks)–which have been launched on iOS and Android phones–will probably catch up first, which will be followed in the next 3-5 years by AR glasses. This is going to be the most successful area of AR/VR. It’s going to be AR first. VR is successful with so many applications within the aerospace, automotive industry, manufacturing, architecture. But it’s not successful on the consumer side at all. There is a gap, we call it the ‘gap of disappointment’. But it will catch up later. What’s going to be a mixture of AR/VR is mixed reality, where you’ll easily be able to slip in and out of a completely immersive VR environment into a partial mixed reality environment. That’s probably going to be successful once we have AR on phones completely integrated with depth-sensing cameras, and we get the machine learning right to identify objects and the reality around us.”
Sylvio was also talking about VR 20 years ago. He was writing whitepapers at the time about the fact that VR will be the future. He was the doing one of the first VR experiences to manage large networking infrastructure in 1993. Five years ago he was talking about the need for a system to assess fake news because he predicted it would be a big problem in the future. To call Sylvio prescient about technology would be an understatement.
“It’s good and it’s bad. It’s good because it’s kind of fun, but it’s bad because nobody takes you seriously all the time, because they don’t see what you see. So it’s hard to get your message across from time to time. But I’m in a place where I’ve worked my way up in Silicon Valley at this point. I’m at a place where I would have to say it’s easier to be heard these days than it was 20 years ago.”