Interview, Fireside Chat, Conference Presentation
Seeing The Future from AI Companions to Personal Software
- The speaker predicts the software category will evolve from a professional developer-led model to a mass-market ecosystem where users create, tweak, and share "mini apps," mirroring YouTube's transition from professional channels to a UGC model, with the expectation that most users will modify existing applications rather than create entirely new ones from scratch.
- Future interactions will center on context-aware discovery where personal apps are suggested based on specific situations (e.g., bedtime, travel) or found through social graphs, allowing for seamless sharing, commenting, and remixing, eventually functioning as "community starters" for niche interests.
- The vision includes a shift away from long-tail, ad-heavy applications toward hyper-personalized, on-the-fly versions built on platforms like Wabi, with plans for users to republish tweaked versions to a mini-app store and for AI to facilitate instant app creation via links rather than manual prompts.
- Hardware and interface strategies are expected to pivot from voice-first or screenless devices to "screen first" smartphones that integrate AI directly into the operating system, enabling local model execution, deep personalization, and proactive notifications that visual interfaces are better equipped to handle.
- The platform aims to introduce visual controls for "vibe" aesthetic customization without technical jargon and features a major update establishing a social graph for app discovery, engagement, and request-based modifications, while ensuring data security through guardrails for sensitive information.
- Monetization and creator dynamics are projected to expand beyond traditional courses, with influencers and professional creators able to publish free software, track authentic usage metrics, and generate revenue through "content, community, and conversation starters" that operate outside traditional app store business models.
- Long-term technical expectations involve apps passing context and memory between one another to break "walled gardens," the emergence of "weird and fascinating" applications that lack traditional business viability, and a hardware evolution toward "truly AI first" devices capable of on-the-go software creation.
- Strategic risks include the potential for severe market consequences due to a lack of aggressive capital investment in generative language models, a missed "generational chance" by shifting focus to video games, and the recognition that the current environment resembles the early internet's "GeoCities" phase before consolidating into centralized platforms.
- The transition is anticipated to begin with simple utilities or "toys" before scaling into a larger platform, with the speaker asserting that this shift will feel intuitive given existing user familiarity with app graphics, though it will be driven by the evolution of AI from simple algorithms to "zero shot" or "few shot" task handling capabilities.