Watson's victory over two Jeopardy champs last week isn't just the first step in human domination by machines, it could also be a step toward better, more interesting video games.
We spoke with Chris Butcher, the engineering lead in charge of establishing the future technology direction for Bungie, about what all of this could mean to gamers.
Here are his thoughts:
Kotaku: What do you think Watson's victory in Jeopardy means to humanity and technology?
Butcher: There's been a lot of media coverage comparing Watson to "Skynet," with the implication that he represents a step towards humanity's enslavement by technology. But if you look at the capabilities that Watson actually has, that argument doesn't hold much water. How exactly are we going to be enslaved by a computer through its ability to answer "Who was the 5th president of the Seychelles?"
In fact Watson is the most recent example of humanity's empowerment by technology. The scientists at IBM have created a great suite of tools for humans to use. By attaching a question-answering engine to a huge pattern-matching information database, they've demonstrated a proof of concept for machines that will be ubiquitous in our lives in ten years' time. Currently, Watson's databases are custom formatted and stored locally. But as these algorithms get utilized more widely for distributed information storage, we'll see the emergence of the Semantic Web where anyone can ask these questions of the entire Internet. Soon, we'll all have our own Watson. And that will unlock a huge amount of human potential.
Kotaku: Do you think Watson has any practical use in gaming? If so what would it be?
Butcher: Natural language processing and question engines have two main applications in gaming – firstly by simplifying the interface. Games (and user experiences in general) will start to be able to interact with us in our own language. The combination of NLP, question engines and gestural interfaces will allow us to interact with synthetic experiences as if they were extensions of the real world. There will still be action games that use custom controllers, they likely aren't ever going away.
But even after NLP starts to make inroads into the interface, characters that use natural language will take a long time to show up in games. Think about how long we've been struggling with the uncanny valley of graphical rendering for human faces. We haven't really begun dealing with the equivalent problem in AI, of creating synthetic characters that use and respond to natural language in an authentic fashion.
Kotaku: How important is AI to good games?
Butcher: Great AI responds to the player in ways that are expected but not perfectly predicted. It's a subtle dance, requiring AI to sense the player's intentions and actions, and work together with them to create a unique experience every time. I love being surprised by AI, it can create a ton of interest and replayability. Without that element of surprise and improvisation, a game can feel like a sequence of canned pre-authored experiences. That turns it into one-time consumable content, a challenge to be beaten and then put on the shelf.
Kotaku: Computers have beaten the best of us at chess and now trivia and it was a big deal. But they've been beating us in games like Halo Reach for years. I sense there's a difference though. Is there?
Butcher: The most interesting part of Watson is the novel pattern recognition and language processing algorithms that IBM's scientists created. But the part that nobody's talking about is the buzzer timing. It's widely recognized in Jeopardy that buzzer timing is a huge part of the game. The human contestants knew a lot of the same answers that Watson got, but they were unable to beat it on the buzzer. So the Jeopardy victory comes from combining some really impressive AI algorithms and some totally unfair timing advantages. The same is true for all action game AI. It's trivial to make a bot that can out-react and out-aim a human player. The challenge of game AI comes from the soft problems of strategy, believability and fun – issues that Watson doesn't have to deal with.