When it comes to future of how our brains will control (or be controlled by) video games, there’s only one thing Valve knows for certain: it wants to measure as many minds as possible.

Mike Ambinder, Valve’s Principal Experimental Psychologist, gave a talk at the Game Developers Conference on Friday that outlined what he believes are the many possible futures of the way that we play in terms of Brain-Computer Interface (BCI). Currently, we mere humans playing video games have an intent, then use physical actions (like button presses or head moving) to interact with the game and the game responds. BCI explores what would be possible if the human interaction was eliminated from this clumsy pipeline, essentially turning intent into gameplay or vice versa.

“How will you interact with technology in 10 years?” he asked, as he laid out many possibilities for how BCI could be implemented, measured and used by game designers to create enhanced capabilities in their titles. “If you think about having a conversation, there are verbal and nonverbal parts. When you get those parts together you get a complete conversation. We might be missing the nonverbal parts of the conversation.”

Beyond the most common interpretation of using BCI to control what’s happening in a game, he talked about seeing a future where BCI could continually measure players, where games could adapt to the players’ cognitive states and even where games could affect those states directly. He also referred to “The Matrix” a number of times.

Ambinder’s talk was almost entirely hypothetical. Though he detailed some of the tools that Valve is using to measure these brain and thought patterns, he gave very few actual innovations or information about gaming interface changes that could come in the near or far future. Instead, he asked question after question about what could be possibly be achieved through speculative data, working towards theoretical benefits.

The most concrete implication for present and future BCI use, which he said Valve is currently using, is in play testing.

“What happens when you get additional information from the player that we’re not currently getting from data input?” Ambinder asked. “We would get an improved player experience If we had access to internal states and emotions and cognizents.”

Again and again, he called for the collection of vast measurements of player data, so developers can assess what a player is feeling when they play a game. He stressed the need for using that data to determine how to design a game to maximize the emotions that a developer wished to inspire.

“We simply want to understand locations of increased levels in the brain,” he said. “If we could take these patterns of activity and translate them to experiences, that’s the key.”

In Valve, Ambinder said they were using a variety of different tools and machines to take the measurements of players to begin collecting this data. Those tools include an electroencephalogram (EEG) headset and devices that measure heart rate, skin response, eye tracking, facial expressions and more.

Though the data, he said, “is noisy.” Which is why he believes that it should be massively collected, so researchers, developers, and assumedly others, can find trends in what the brain is experiencing.

“If we could get the advantages of direct observation and internal states, we’d be much better game designers,” Ambinder said. “We could always be playtesting on everyone.”

Valve, as a hardware development company, has been no stranger to experimenting with possible interface changes, or pushing forward the expectations of how players can interact with games. They also created the Steam controller, an unconventional input device for playing games that replaced controlling analog sticks with trackpads. Valve is also responsible for developing the Vive virtual reality headset, one of the leading technologies in the VR space.

As a possible way collect and track this data, Ambinder talked about putting sensors like the EEG in VR and AR headsets as a way to gather those measurements on a large scale.

“One advantage of VR and AR is you’re getting consistent contact with brain activity, so you might be able to do some interesting things,” he said.

One of the larger benefits of having these measurements, as he saw it, is through the possibilities of creating adaptive gameplay, where the game actually changes based on what the player was experiencing.

“Game dynamics would be no longer static if we had access to this data,” Ambinder said. “Just as players respond to the game, the games could begin to respond to the player. As a consequence, gameplay becomes adaptive and personalized.”

Essentially, he looked as collecting that brain pattern data and using it to deliver the type of game or game experiences that would attract players the most. He talked about measuring his own thought patterns while playing Valve’s “Team Fortress 2,” and seeing what happened to his brain waves when he collected a rare item or killed an opponent. The maximization of those sensations through gameplay is one of the possible eventualities of BCI as he sees it.

“Difficulty levels would become a thing of the past,” he said. “We’d have dynamic difficulty adjustment. Difficulty that’s designed for you.”

And then there is the method for crafting that interface between the brain and the computer. Ambinder said that the future of this technology could be non-invasive, like the EEG headset, or even invasive technologies like implants.

Which brings us to “The Matrix,” the completely simulated world featured in the late ‘90s sci-fi movie, which Ambinder discussed as being almost a foregone conclusion. He said all of this work, stimulating synaptic responses in order to elicit desired emotions or reactions, essentially ended in matrix-like experiences.

“It could be decades or centuries away,” he said. “But the path to that reality, people can see it and they’re actively working on it.”

However, Ambinder continuously maintained how hazy the future is about how BCI could evolve and whether all or any of what he considered in the talk would come to fruition.

“Neuroscience has a long way to go,” he said. “There’s so much we don’t understand. Like, how do you think a thought? How do you play a game.”

Of course, he would not give away Valve’s research secrets in a GDC talk. So, it’s entirely possible that the industry giant is crafting marketable tools to measure brain states, create adaptive games or to take us closer making “The Matrix” a reality. But for his presentation, Ambinder kept the whole topic extremely abstract.

“I’m optimistic, but we don’t know yet, we need to create something and see,” he said about some theoretical technology or implementation. “It can’t just be theory. We have to make something that actually works.”