This year’s GamesBeat Summit 2025 teemed with developers, creators and industry leaders. They explored the exciting possibilities that artificial intelligence (AI) holds to revolutionize game development. The program featured an extraordinary roster of experts and thought leaders. Other prominent speakers were Chris Melissinos, principal evangelist for video games and immersive technologies – AWS, Aaron Farr, co-founder and CTO – Jam & Tea Studios, Ashwin Raghuraman, a senior solutions architect for games – AWS and Hilary Mason, co-founder/CEO of Hidden Door.
The summit convened in San Francisco on March 15-16th. Its major thrust was requiring generative AI tools to be built into games, including a heavy focus on transformer models, large language models, and image generation models. The goal of each of those discussions was to showcase how these technologies can enhance and accelerate human creativity. They also acknowledged the gaps that remain within the industry.
Embracing New Tools
Chris Melissinos, artist and game designer, opened the discussion with the panelists, focusing on the creative potential of AI in gaming. He illustrated the ways in which AI was able to automate many of the core functions artists and designers have always done.
“It’s automating, in some ways, the scaffolding that allows an artist, a writer, somebody doing a level design, to be able to then come into something without having to build all the pieces from the bottom up.” – Chris Melissinos
Melissinos was adamant about their importance in treating generative AI as an assistive technology. He thinks it should complement, not supplant, human creativity. Firstly, he underscored that AI holds tremendous potential to increase production and productivity. As an author, you need to be careful of your own bias when judging AI-created work.
Aaron Farr, too, expressed the same appreciation. He stressed that even though new technologies are cool and shiny, new technologies can’t wipe out old tools just yet.
“I have a repertoire of tools. What does this add to it? Where does it fill in gaps? I’m not throwing out all the old tools. I need a return on investment,” – Aaron Farr.
Farr warned that innovators need to approach AI mindfully to ensure they embrace it without succumbing to “premature optimization.”
The Role of Generative AI
Throughout the course of the summit, these speakers went further in-depth on the unique AI models already transforming the gaming industry. Ashwin Raghuraman discussed the opportunities LLMs and diffusion image generation models have created. So he focused on their potential to help with brainstorming and conceptualizing ideas. This technology opens the door for exploration and helps developers to ask more cutting edge questions about their projects.
On the topic of what’s hard about generative AI, Hilary Mason shared some of her experiences to illustrate that generative AI is not the case of one big red ‘make me a game’ button. Rather, it creates thrilling new opportunities for innovation.
“It is not a big red ‘make a game’ button, exactly, but it is something that then lets us ask questions.” – Hilary Mason
Mason added that the development, monetization, and distribution of game mechanics are changing so quickly that an agile approach to development is essential.
“Our cost functions have changed, our capability functions have changed. What we can let our players do has changed,” – Hilary Mason.
That flexibility is key to unlocking the true benefits of generative AI to bring never-before-seen experiences to life in games.
Future Considerations
That conversation at GamesBeat Summit 2025 looked beyond the impacts of current AI technologies to envision where the entire game development pipeline is headed. Aaron Farr acknowledged the current limitations in fully automated game creation but expressed optimism about leveraging new technologies to create novel experiences.
“We’re trying to apply this technology in ways that create something new and novel that we couldn’t do before,” – Aaron Farr.
Specifically, he emphasized the need to embrace both the promise and peril of technology. This means looking into surprising uses to push the development of anti-harassment/abuse practices into more innovative directions.
Farr’s insights, though, brought us back to the incredible potential of generative AI. The onus is on developers to remain vigilant in addressing these issues and biases.