Just to recap, on Friday, February 20th, Microsoft announced that Phil Spencer was retiring and
Asha Sharma would be the new head of Microsoft Gaming. Prior to her new role, Asha was most recently head of Microsoft's CoreAI division, tasked with putting Copilot in as many places as possible.
Now she's Xbox's new CEO. Coincidence? Not on your life.
Microsoft is sending a very clear message about the future of their gaming endeavors, and while we don't know exactly how it will play out, AI is certain to have a front row seat.
That's the only reason they would put a non-gamer, AI executive, in the captain's chair. There is no other reason.
Just to level set, if it needs to be said, being CEO of Microsoft Gaming is a big deal. It's a top CEO job, and probably highly sought after. You don't get a job like that by mistake. There were a ton of qualified candidates. Microsoft could have even chosen to poach someone from Sony if they thought PlayStation had a winning playbook. Or they could have elevated someone within the current Xbox team. Or they could have hired from any number of places for that role. Believe me, almost anyone would have jumped at the chance. But they gave it to someone who isn't from gaming (that's clue number 1) and someone who is all about AI (that's clue number 2). Cyberdyne's (oh sorry, Microsoft's) message couldn't be clearer.
.png) |
| I'm not a doomer. I just like to make fun of AI using images from Terminator. |
AI will be an integral part of the next generation of Microsoft gaming.
This could be the reason we see a new Xbox sooner rather than later from Microsoft. It didn't integrate AI into the hardware of the Series consoles, but you can bet they will in the next generation, whether we get a meaningful boost in power for gaming or not.
But this isn't really news for anyone who's been paying attention to what's been happening in tech and AI in general.
AI is the next frontier and every tech company is chasing it. Just last year, Microsoft invested roughly $80 billion into data centers. Just to put that into perspective, the Activision/Blizzard acquisition was $68.7 billion. Where do you think their priorities are?
So AI is a given. It's coming whether we like it or not, and A.I.sha is the perfect CEO to cram AI into the next generation of Xbox consoles and the Microsoft Gaming ecosystem.
But it was coming no matter what, and I don't say this meaning to throw any shade at Asha. She might do an amazing job and truly help usher in the next great generation of Xbox. But trying to pretend AI won't be front and center in the next generation is silly. That would have happened regardless of who the next CEO was.
So how could A.I. being integrated into Microsoft Gaming and the next generation of Xbox hardware look like?
I have some thoughts...
I can see at least 3 key areas where AI could realistically be utilized in the next console generation:
- Game development - traditional
- Game engine rendering
- Game dialogue generation
What do I mean? Let's briefly dive into each one.
A.I. in traditional Game Development
Like it or not, AI can be used in a variety of ways to speed along asset development for traditional games (I'll get into the distinction of what is tradition versus not in a moment). It can be used to:
- generate textures and select in-game scenery
- write draft outlines of certain game scenarios
- refine drafts of game dialog
- write game dialog
- generate certain sound effects and select spoken in-game dialog
This of it like A.I. being used to generate all the background, NPC, not core assets used in a video game. For the foreseeable future I see core, important aspects of game development staying human-generated. For example:
- Core game character dialog and voice acting
- Core game character models and textures
- Core environment design and texture mapping
- Overall game script and roadmap
Basically, all the most important aspects of a game, the things core to the overall experience, will likely continue to be human made for the next generation, but all the ancillary things that are kind of just in the background or used to fill out a game environment, but wouldn't have had a lot of time spent on them anyways, will start to be more and more AI generated.
And sadly, this will result in work-force reductions. Not because they have to happen, or because the people don't have value to add in other parts of a game's production, or that they could then spend even more time on the human-generated parts of the games, but they won't. There is a general expectation that the introduction of A.I. will result in workforce reductions, like there is an expectation when two companies join together that there will be "synergies" and work force reductions. Do there have to be? No. But there will be.
A.I. in Game Engine Rendering
Another area I can see A.I. immediately helping with is direct rendering of environments and assets inside of a game engine, on the fly. Imagine a game like Minecraft, that is already procedurally generated. It creates new game worlds each time you begin, based on rules that its designers created. Now, what if, you had a game like that, but instead of the same rulesets being designed in advance, an A.I. engine had a set of parameters to follow and created new game worlds on the fly as you entered them? What if every tree texture in Minecraft had slight variations so that they didn't all look like cookie cutter elements? What if enemies and NPCs had slight variations so that no two looked identical, but it all mostly still looked like the Minecraft you know and love. A.I. can basically help with that today, and I'm assuming if you built A.I. into the hardware itself, that could run some of these game engine renderings locally, although you might not need to.
Now imagine this in a game where an entire library of base game assets had been created by a human designer, and now A.I. simply iterates on what has been created to produce variations in the game world. In theory, no two NPCs would need to look the same.
In theory this could be cool, but in practice we'll have to see what kinds of experiences are created.
A.I. in Game Dialog Generation
Similar to the idea of Game Engine Rendering, game dialog could also be rendered by A.I. real-time as you interact with an environment. Ideally, human writers would create a base script and a lot of samples of dialog for the A.I. to build off of, but then while you are playing the game, it would create unique dialog based on the character you are interacting with and the evens that have occurred in the game so far. This could truly create a different experience for every person playing a game, and create a different experience every time you went back to play a game. Sure, the story would be the same overall, but the unique aspects of character dialog would be slightly different each time.
Maybe you think this sounds great, or maybe you think it sounds awful and terrifying, but I can definitely see it coming.
.png)
I'm sure there are a lot of other areas A.I. could be used in game creation and generation, but based on my own experience, these are the areas I can practically see it working in the near term.
Ideally, all of this will lead to shorter development cycles, lower costs, and faster game releases. This is practically an expectation, and I'm sure Microsoft Gaming will do everything in its ability to ensure it happens.
So in the end, A.I. is coming to Xbox whether we want it or not. There will be more job destruction along the way, and likely some slop we'll have to put up with in the near term. We'll just have to see exactly how it all plays out.
.png)
.png)
.png)
.png)
.png)
.png)
.png)
.png)
.png)