As AI costs rise, there is little evidence of major utility
The article argues that despite rising investments in AI tools for game development, there remains scant evidence of significant practical utility or cost-benefit justification for their widespread adoption in the industry so far.
Background
- This opinion piece argues that generative AI (tools like ChatGPT, image/video generators) has yet to show significant practical value in game development, despite heavy industry investment and hype.
- Major studios (Ubisoft, EA, Microsoft, Riot, Roblox) have been touting AI as a way to cut costs or speed up production. Some have laid off thousands of workers, partly citing efficiency gains from automation.
- The author claims the actual utility in game dev remains narrow (e.g., early concept art, placeholder assets, some code snippets) while costs—compute power, licensing, energy—keep rising.
- Running and training large AI models is extremely expensive (e.g., OpenAI reportedly loses money on ChatGPT despite subscription fees), raising questions about whether the technology can deliver a return in a medium where polish and originality still require human craft.