Building AI for consumer applications isn’t all fun and games

Key takeaways

The integration of AI in consumer applications, especially in gaming, must prioritize user experience over the novelty of interacting with AI.
Universal mutability in AI allows for real-time adaptations to user preferences and context, enhancing gameplay and interactions.
Cost optimization remains a critical challenge for deploying AI models, yet innovative approaches can significantly reduce expenses while preserving quality.
Successful AI applications facilitate human connections and enhance experiences rather than replacing them.
Developers must focus on creating applications that evolve with user needs to keep pace with changing consumer expectations.

Summary

The Journey into AI and InWorld

Kyle Gibbs shares his journey into the world of AI, starting from his consulting role where he developed machine learning practices, to his foundational work at DeepMind. He details how he transitioned from theoretical research to applying large language models (LLMs) into real-world applications at Google. This experience led him to co-found InWorld, where he and his team focus on making generative AI accessible to consumers across various applications beyond enterprise use.

Navigating Technical Challenges of AI in Gaming

Gibbs discusses the integration of AI within consumer applications, particularly in gaming. He explains that while AI can enhance user experiences in games through real-time interactions, it poses challenges such as natural dialogue generation and ensuring coherent back-and-forth exchanges with characters. Instead of focusing solely on AI interactions, the aim is to facilitate user experiences through AI's ability to support gameplay and social connections.

The Role of User Experience in AI Design

The podcast emphasizes the importance of user experience in the design and deployment of AI applications. Gibbs asserts that successful AI integrations should enhance real user interactions rather than be the main attraction. He highlights a key insight: interactions with AI should serve to facilitate engagement with beloved worlds or individuals rather than being self-referential.

The Future of AI in Consumer Applications

Looking ahead, Gibbs shares insights on the potential for AI to personalize user experiences in games and applications. He discusses the idea of universal mutability, where AI can adapt itself in real time based on a user's context and preferences. The goal is to evolve applications continuously while focusing on user needs, ensuring that AI aids rather than overwhelms the user experience.

Cost and Scalability of AI Models

A significant portion of the conversation is dedicated to the cost associated with running AI models and the challenges faced in scaling them economically for consumer applications. Gibbs discusses how they managed to optimize costs dramatically while maintaining quality, showcasing their innovative approach to AI through the development of their speech model, TTS1.

Show notes

Kylan Gibbs, CEO of Inworld, joins the show to discuss the technical challenges of creating interactive AI for virtual worlds and games, the significance of user experience, and the importance of accessibility and cost-efficiency in deploying AI models.

Episode notes:

Inworld provides solutions for AI applications that allow teams to build and deploy workloads, spend less time on maintenance, and accelerate iteration speed.

Connect with Kylan on LinkedIn.

Today we’re shouting out the winner of an Illuminator badge, user MrWhite, who edited and answered 500 questions, both actions within 12 hours.

See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.