The AI Revolution in Game Development
Exploring how AI-powered tools are redefining game development, enabling indie studios with superior design, coding, and level creation.
Hi there!
This week on Eximius Echo, we explore how AI-powered game development is transforming the industry. By enhancing design, streamlining coding, and automating level building, AI empowers indie studios to create innovative and immersive games. Though still emerging, this technology holds great promise for reshaping the gaming landscape and levelling the playing field for smaller developers.
If you’ve not heard of us, Eximius is a pre-seed stage fund focusing on FinTech, SaaS, Online Media & Gaming, and HealthTech. You can find out more here. This newsletter is an attempt to share ideas, insights, and context within the realms of our chosen sectors. Let’s dive in.
The gaming industry is a content behemoth that produces tens of thousands of games every year. These tend to range widely in terms of graphical fidelity, world design, gameplay, etc., which makes every experience unique. While this has always been a treat for gamers worldwide, new-found improvement in gaming infrastructure is all set to drastically reshape this experience forever.
A big reason for this wave of adoption is going to be indie studios. Just taking Steam as an illustration, there were roughly 14,000 games released on the platform in 2023 with indie studios releasing ~13,800 of them (98.7%). Aside from creating more content, “indies” also look to be more expeditious with content development. 54% of indie studios conclude their prototyping windows within 3 months, while 39% do the same in under a month. While larger studios have in-house tech that already enables them to create industry-leading games, indie studios have struggled to make games of that scale. AI-powered tools have the potential to bridge that gap.
This leads us to a key question – ‘How exactly will AI be used across the game creation value chain?’
The answer to this is pretty multi-faceted as indicated above. We will explore the top 3 use cases and understand how exactly AI is altering the landscape.
1. Character-work
Designing characters is one of the most technically complex parts of game development. While streamlining animation is the biggest problem statement, developers (specifically indie studios) are also looking for ways to improve upon their process of generating character ideas and making them seem as life-like as possible.
Over the past year, AI has become especially proficient in generating models, assets, and NPC behaviour mapping. Shape-E, as an illustration, is a model created by Open AI that can generate 3D models through text or visual inputs. AI is also more proficient in tracing the real-life action frame of the characters frame by frame which can make animations seem more life-like and organic. In terms of NPC behaviour mapping, AI can bring autonomous agents that can be programmed to act on their own accord. Thus, players can speak to NPCs using text and/or audio to engage with them and get realistic answers. Nvidia also demonstrated a similar project with Inworld Engine which showcased this future of character interaction.
Therefore, AI can not only help studios generate character art more quickly, but it can also ensure that NPCs can realistically fill the game’s world.
2. Code Writing
Game development takes an excruciating amount of time and resources. While it is possible to develop fully functioning games in months, larger and more complex projects can end up taking several years due to their intricacy.
One of AI’s most popular use cases is in generating codes from scratch and assisting developers to make them more efficient. As per a report from Statista, 85% of developers have used an AI assistant for generating code. Another from Github, the parent company of Copilot (an AI-based code completion tool), revealed that the tool was used to generate over 82 billion lines of code in its first year alone.
It is important to note that this technology is still in a nascent position. Therefore, while we can expect AI to lift a large chunk of the developers’ burden to enable even smaller studios to tackle larger projects, there is still considerable progress yet to be made on the technology.
3. Level-Building
This has a two-fold application: generating art for the game’s world and generating individual game levels. Let’s address both in detail:
Tackling the former use-case first, tools like DALL-E and Midjourney allow users to generate novel art through textual prompts in a seamless way while also allowing them to iterate on the output to have it be as precise as possible. Therefore, they aid developers in not only ideating on design but also execution. Entire game worlds can be created taking the assistance of these tools. While more than 2 million images are being created daily on DALL-E 2, Midjourney has been able to amass over 16 million users so far.
Taking note of their utility, quite a few popular studios have begun incorporating AI art for their upcoming projects. Square Enix, for instance, admitted to using Midjourney to generate art for their upcoming game Foamstars. Even Treyarch, the developer behind major Call of Duty games, is reported to be using AI art for future instalments. This is expected to help studios cut down ideation and prototyping costs and lend significant aid in generating new ideas.
The second application lies in not designing game worlds, but individual levels. This is specifically useful for hyper-casual and casual games that feature a consistent level design. Here, AI can greatly amplify the efficiency of procedural content generation to automate the creation of new maps and levels. This allows developers to maintain a consistently well-stocked inventory of levels that can be introduced to the game without breaking the game design cohesion.
AI can also be used to dynamically scale the difficulty quotient of levels to ensure a more consistent design curve. This can be integrated into level design to keep the players hooked to the game for longer periods and reduce churn.
We already see popular new-age games like No Man’s Sky, Starfield, and Hades use procedural generation to generate new game levels.
Closing Thoughts
As indicated above, AI has the potential to be a great leveller for the industry. While larger studios were able to accomplish much of what was discussed above at a greater scale, indie studios and start-ups were always dealt a shorthand. This technology, on the other hand, grants smaller studios the capability to target more complex projects.
This trend is being seen in Indian start-ups as well. Quite a few prominent new-age companies, like Triple Tap Games, are looking to leverage AI heavily to optimise their art design and generate new levels at a quicker rate. Others like Second Quest are building a layer of modularity in their games to both generate content quicker and allow players unprecedented ease in UGC (user-generated content). Even larger entities like Jetapault are building a large chunk of their new projects using generative AI and existing assets.
This is why we at Eximius Ventures are so excited to see what new start-ups are going to build with this technology. If you're an entrepreneur building in this space, we would love to hear from you. Please write to us at pitches@eximiusvc.com with your ideas.