The Near Future (2022-2035)

- If the post 7th generation cadence of releases continues, the 10th generation of video game consoles will emerge around 2027-2029, defying multiple past predictions that console gaming would be undone by smartphones and streaming. These machines should be expected to have at least 50 teraflops of GPU performance, or at least the Microsoft and Sony machines should. Nintendo dances to its own tune and is predictably unpredictable.

For reference, Tim Sweeney believes that 40 teraflops is enough to create photorealistic graphics. Ray Kurzweil believes that "Virtual Reality" graphics will be "indistinguishable from real reality" by the end of the 2020s.

The key trends of the 10th generation will be ray tracing, generative AI and the streamlined asset creation provided by the Unreal Engine 5 and competing game engines.

- Ray tracing is already being employed in today's games, but the hardware limitations of 9th generation hardware prevent the effect from being used as a complete lighting solution. It will be common for 10th generation titles to use ray traced shadows, global illumination, ambient occlusion and reflections all at once, bringing a new level of realism to titles while making game development much more simple for level designers. No longer will level artists have to spend time using baked lighting. Every scene will be capable of being rendered under any time of day or weather condition without the extra work to make the scene look correct.

- New graphics engines like the Unreal Engine 5 also promise to further streamline game development through features like Nanite and Lumen. Nanite allows game artists to save time by not having to design different versions of an asset for different levels of detail. Instead, the artist will design one high quality asset and the engine itself will dynamically simplify the asset the further away it gets from the player camera. Graphics pop-in will become a thing of the past because of this solution.

Lumen will allow games to get most of the visual benefits of ray tracing, while not being as hardware intensive as actual hardware-based ray tracing. It may eventually become a moot point as consoles and PCs become capable of opting for the real thing, but it and similar solutions will be useful for the next decade of game development.

- Generative AI needs no introduction for most of you but for those just tuning in: it's the big one. Generative AI will be one of the biggest, if not THE biggest technology story of the decade, potentially bringing as much disruption to our lives as the smartphone, streaming media or the internet itself, if not more so. The same generative AIs that are now disrupting the world of art are coming for everything else. 2D art, 3D art, video, music, writing, voice work. Any entertainment that you can consume on a digital device will be affected, and no industry will feel this as much as the gaming industry. That is because making a video game is a multidisciplinary job. You need coders, you need level designers, you need writers, artists, musicians, actors. None of the positions that I just named are safe from generative AI.

You don't have to believe me, it's already happening. It won't feel very disruptive for the rest of the 9th generation, as game studios will likely be cautious with their implementations, not wanting to attract any of the blowback that's hitting the companies embracing AI art. You'll have some independent developers playing with different AIs to solve some problems, some major studios using it at the periphery. It's usually with new gaming generations when developers decide to take bigger risks on projects, and the 10th generation is when I expect this trend to explode.

The best video games of 2030 will be made collaboratively between humans and generative AIs. Humans will mostly be working on the fun things, while the drudgery gets passed on to the AIs. The player character and plot-important NPCs will be designed by people, while the unimportant pedestrians will be given to AI. The main theme and musical score for key scenes will be composed by people, the ambient filler music will be given to AI. People will voice the story characters and AI will voice the random NPC blurbs. Key locations will be hand designed by artists and level designers, unimportant forests and random city blocks will be given to AI, and so on.

Because of all this, video games will become bigger, more detailed and more realistic. At the same time they will become easier to make than they ever were before. Small, independent developers who had to settle for sticking with retro 2D or highly stylized aesthetics will now be able to choose to make their own massive, 3D open worlds. Large, established studios that are already at the AAA level can choose to put out the same quality of work with less human employees, or keep them and do even more with the added abilities for AI. I think (hope?) that ongoing competition in the market will drive employee headcounts to stay steady. The company that fires employees would be at a disadvantage versus the one that keeps them, and the fired employees might just become the old company's future competition. Why not? AI gives them the tools to do so.

- VR Headsets are here to stay and will continue to move forward, iterating towards smaller, cheaper and more powerful headsets. We will go from having to slide these big, awkward bricks over our faces to using a device that is no more obtrusive than a pair of bulky sunglasses. It will still be too early for VR headsets to displace consoles, as the couch and game pad experience will remain attractive to people who don't want to physically exert themselves in VR. However, I wouldn't be surprised to see Microsoft introduce its own gaming focused, Xbox branded VR headset to the market in his time frame. Apple likewise.