What Other Industries Can Learn from Video Game Innovation

Artificial intelligence is becoming increasingly mainstream, with more businesses committing budgets and teams to AI-enabled processes.

But for the video game industry, AI has been a mainstay for several years in varying degrees of complexity.

From simple algorithms used to control the behavior of non-playable characters to intricate systems powering the development of virtual worlds, AI has powered gaming for the past few decades.

What can other industries learn from the video game sector in terms of approaching transformative innovation?

AI Business spoke with Yves Jacquier, executive director at Ubisoft Montreal, about how his company’s approach to innovation has developed since he joined in 2004. Yves will speak about transformation innovation at the AI Summit in New York.

The following is an edited transcript of the interview.

To catch a glimpse of the conversation, watch the video.

AI Business: In terms of transformative innovation, how far has Ubisoft come since you joined in 2004?

Yves Jacquier: The way we made the games in 2004 has changed a lot over the past 18 years. In 2004, a big game was made by tens of people collocated in the same studio. Today, a big game requires hundreds of people often located in different studios. Since 2004, we've been through many different transformations. For example, we have honed core development expertise with a dedicated team to enable this. Prior to the pandemic, we were already able to work on a single game from different studios, so we were able to support work from home within a few weeks when COVID hit.

When you want to assess the quality of a game, you need to ask the player. But this has limits. When you have very complex games with different systems interacting together, you want to measure if the game is well-balanced as a whole. We introduced telemetry which is measuring what's happening in the game with the objective to get better insights into the way the game is balanced and whether it keeps being fun. This has changed a lot in our operations, but also in terms of skills that were acquired by the teams to make the value of that. And these are only a few examples, as we as an industry have been at the edge of many of the technical revolutions, such as AI and machine learning.

AI Business: When an idea just works – and works well, case in point, the original Assassin’s Creed, how important is it to build on that initial breakthrough at pace?

Jacquier: Making games is extremely difficult because you need to align artists, programmers and designers into one single direction with the objective to enrich the player's life by providing fun experiences. In terms of pacing, it's a question of brand management. And the core question is what is the promise to your players? In other words, what should be familiar from one experience to the other? And what should be surprising, totally fresh and new, what should be enhanced from one experience to the other?

AI Business: How much should user base needs and demands come into question when considering potential innovations? For example, Ubisoft’s Zero Harm in Comms project.

Jacquier: The objective of the Zero Harm in Comms project is to start tackling the issue of toxicity of in-game chats. We have been working on this topic for a long time along with our competitor Riot Games. Both companies reached the same conclusion that currently off-the-shelf tools do not cover our needs and that we need to align to solve the problem. The idea here is to be able to share data between competitors but to do it in a way that preserves privacy and confidentiality so that we can train AI together to better detect harmful content in chats. 

We all know and have to acknowledge that online toxicity is an issue. It's not a Ubisoft issue. Nor is it a Riot issue. It's an industry-wide issue. It goes beyond social networks. It's the same on all online channels. By trying to align on those topics we feel that we can provide prompt answers. 

The most important part is truly being connected with what's happening outside with the users with the way they're using your products and the experiences that they're having because they are able to tell you loud and clear that they want an innovation to help them solve a problem or to enhance an experience. 

In the case of animation, at Ubisoft La Forge, we came up with new concepts and ideas that were totally new within the gaming industry. Those ideas helped unlock new possibilities in terms of immersion or diversity of characters. That was simply not possible with traditional methods. When the first iPhone was unveiled. Nobody asked for it. We all had flip phones at the time used solely to make phone calls. Now, we barely make phone calls on my cell phone, but we use it to purchase tickets or order from the grocery store. What this means is that you need to be connected with innovation to solve short-term problems and make sure that you can add value compared to what exists already to solve this problem. But also try to anticipate not only what the use cases of tomorrow will be, but also the kind of issue that you can have in the future to try to pave the way to solve them along the line.

AI Business: As an R&D-focused team member, how frustrating can it be if your team does the necessary work, but only for budget cuts or other factors to impact your initial vision of the work?

Jacquier: When COVID hit us, we were in the middle of the production of Assassin's Creed Valhalla, and suddenly lockdown. So how do you end the production of such a giant game like Valhalla, especially when it comes to recording the actors to finalize the last cinematics, for example? 

For us, it was also an opportunity to push forward some technologies that we have developed, such as text-to-speech or voice conversion prototypes, so that we were able to do some last-minute retakes without going in the way of the production and with the agreement of the different actors involved. It’s really trying to take a step back on a situation and making sure that one way or another, you will learn from it and you adapt in front of it.

AI Business: Based on your experiences developing Valhalla that it improved your team's ability to work going forward onto future projects?

Jacquier: It was difficult. There were many challenges. However, it was also an occasion to improve things, to reinvent some of our recipes and find very valuable use cases that were probably not that obvious before. It's really a matter of posture and mindset as well. When you're curious about things, when you're trying to solve problems creatively, when you're trying to go beyond the silos, I'm pretty optimistic that you can improve something out of that. Wherever it's the shared knowledge that we have, what works and what does not work, but also concretely showcasing some creative ways to solve problems and to create values.

AI Business: There’s a lot of industry talk about teams working together holistically to obtain better results. How exactly can brands implement this concept effectively?

Jacquier: When we talk about innovation, we're talking essentially about a mindset about culture and this goes up to a cooperator culture. When we created La Forge, for example, we came up with the idea that silos kill innovation, in the sense that disruptive ideas come when you mix diverse disciplines, experiences and point of views. When you have experts in one field innovate in their fields, it leads to incremental innovations, which is great. It means improving the existing entities. 

When it comes to disruptive innovation comes questions that you can't predict. If you want to create a system that generates animations, only an artist can tell what is really important. If you want to create a system that targets harmful content, not only do you need specialists in AI, but you need also to bring experts of other fields like social science or communications, but also different sensitivities to what harmful content is. And all of this raises many questions. Often there is no clear legal frame, for example, and you need, from a corporate level or from a team level, to create your own policies such as data policy. And to do it the right way, this must be enriched with diverse expertise and point of view as well to define what's fair, and then do what's fair. 

For us, breaking those silos and making sure that by design we collectively co-create our innovations is really a way to ensure that we diminish our blind spots and that collectively, we're able to address the most important questions.

We had one project that involved data from our employees from an HR standpoint, and from a legal standpoint, everything was fine. We could use it. However, at some point in the process, and I will explain that at the AI Summit, we were not comfortable using this data and we created our own data policy around that. By doing this, we were able to anticipate what came a couple of years later which was the EU's GDPR. And because we took that into account, we anticipated that, and we created our own data policy beyond what was legal at the time. We didn't have to roll back anything. We were practically compliant with GDPR. That's an example where you only can have holistic innovations and make sure that it goes beyond your lab and innovation teams. Only if you're able to provide this holistic approach, holistic expertise, holistic point of view and holistic cultures.

AI Business: What can other industries learn from Ubisoft and the wider video game market in terms of approaching transformative innovation?

Jacquier: Video games are transversal by essence. We create virtual worlds, virtual characters, and rich interactions. We trained virtual autonomous cars in virtual worlds before it was done in real life. There's a field of application first to be found, and different techniques, different challenges, and different concerns. And I see the opportunity to create bridges and learn from different industries and points of view. We learned a lot from the health industry, for example, and they learned a lot from us by creating virtual avatars together. And the same thing goes with social science or climate change, just to name a few. We were able to different industries or different activities able to enrich its offers on very concrete applications by the very nature of what we're doing in our virtual worlds. 

But transformative innovation is less a technical challenge than a human challenge at the end of the day. I will have the chance to explain our take on that at the AI Summit. And these are things that can be transferable in many contexts beyond video games.