Hello everyone! This developed pretty suddenly, but for next week's podcast we're going to have Emily Short as a guest! For those who don't know her, Emily is a writer and game designer specializing in interactive narrative, dialogue, and social interaction modeling. She's also the Character Engine Project Manager at Spirit AI, which is a tech company focused on artificial intelligence within academia and the game industry.
This is going to be a much more structured conversation mostly on human-computer interaction and the possibilities and potential of that within video games. I've already come up with several questions myself (which I've included below) but if any of you have additional questions you'd like to ask feel free to let me know! We'll be recording this podcast on Saturday, so you'll need to submit pretty quick!
My questions were largely inspired by a talk Emily gave at Gamelab Barcelona last year, so if you guys would like to listen to that talk here is a link to it: https://www.youtube.com/watch?v=n9kC61d3fQU
Personally I'm very excited for this discussion, as it's a topic I'm pretty passionate about myself as a fan of strong narrative in video games. Can't wait to see what you guys think! -Mike
My questions:
1. “New algorithms to create new experiences vs. architectures that reliably support specific game designs.” Touch on how R&D on games today focuses a lot more on finding new and exciting ways to kill people rather than on how to make interaction with characters more meaningful. What can be done to change this trend in game design priority? Can complex character interaction be exciting enough to be marketable and profitable as a game’s defining feature?
2. Based on the response you got from that player for your game Galatea, and taking into consideration how it reinforced your desire to connect players with characters based on how they choose to interact with them rather than on conventional drama from books and movies (backstory, good dialogue, setting, etc.), what kind of progress is being made toward making this a reality in large scale games?
3. Let’s talk about Versu, where you have done a lot of research into real-world conversational pragmatics. Could you describe how players reacted when NPCs preferred to speak to one another rather than to the player? Why do you think players felt that this was broken or unrealistic? I personally hate the way Skyrim (and many other AAA games) have NPCs that always have to have something to say to me when I pass them, which leads to a lot of silly repetition. What is the balance between Versu's and Skyrim’s ways of handling NPC interaction with players?
4. In regard to Character Engine—with the demo you described during your Gamelab talk with the robot being interrogated—with AI able to interpret incoming speech and react to it with improvisational or procedurally generated responses, what kind of challenges will writers have to face when creating their stories and directing players? Anyone who has ever been a DM for a D&D campaign knows that despite how much you might prepare, you can never account for all the ways the other players might choose to act, and adjustments to the story sometimes have to be made on the fly. How do you account for this with game design? How do you control for it? In addition, if the AI is choosing how to respond to player text-to-speech for example, how does that change a writer’s art? Theoretically the writer would no longer be writing dialogue for the game because the AI would choose how to respond, so how does that change what a writer does for a video game like this?
5. What is data science for player modeling? How does adjusting the way NPCs speak to players based on their personal playstyles and preferences change the overall experience from one person to another? Do you think these alterations would be significant enough to suggest that two people who bought the same product played a different game?
Paul Metcalf
2018-04-26 17:26:17 +0000 UTCResonant Arc
2018-04-26 17:22:42 +0000 UTCPaul Metcalf
2018-04-26 17:16:49 +0000 UTC