The media landscape has been abuzz with discussion about AI's impact on the operations and processes of content creation and distribution. However, far less airtime has been spent on how AI is transforming the interfaces through which consumers engage content from that pipeline. This shift has profound implications that deserve closer examination.

My Journey with AI-Driven Interfaces

Since early 2024, my experience with tools like Perplexity and ChatGPT has fundamentally altered my approach to information gathering. Unlike traditional search engines, these AI platforms provide direct, comprehensive answers, eliminating the need to sift through multiple links.

A recent family vacation perfectly illustrated the emerging potential of AI interfaces. While searching for a movie to watch, we encountered the all-too-familiar frustration of having to navigate multiple streaming platforms (each with a discrete subset of the content universe) to find something to watch. It’s a modern predicament that Evan Shapiro colorfully and aptly described as follows:

Think about your own experience. How many times have you searched for something to watch on an SVOD. Failed. Logged into another SVOD. Searched, failed. Logged onto another SVOD, searched, failed… Gave up, got stoned, and just watched Friends? This is now the new normal TV User Experience.
Evan Shapiro

Instead of wrestling with that, I asked Perplexity for movie recommendations. By providing some parameters and a couple iterative refinements, the Lakamp family discovered and enjoyed "Point Break" - a small but enlightening moment that crystallized an emerging trend for me.

Key Observations on AI-Driven Interfaces

Three critical themes emerged from my experience:

Precise Intent Expression: AI interfaces allow users to articulate their precise intent, moving beyond algorithmic inference based on past behavior.
Comprehensive Information Aggregation: Rather than to navigate a search across disjointed sources, users can now access synthesized information that relieves the user of significant work.
Paradigm Shift in User Experience: The implication is that traditional apps and websites will be gradually replaced by more intuitive, intelligent interaction models.

Worth noting, I believe those themes will apply broadly to the evolution of interfaces, not just consumer interfaces. That is, they will certainly apply to consumer interfaces, but also to B2B applications, enterprise tooling and beyond.

As an aside, I'm not suggesting that browsing or apps are going away. Both will persist meaningfully. Websites didn't go away once mobile apps came onto the scene.

"AI is the New UI"

Nonetheless, I'm convinced that "AI is the new UI". This phrase, used by Qualcomm CEO Cristiano Amon during his 2025 SxSW presentation, encapsulates a transformative vision. Amon argues that the most significant upcoming change lies in how we'll interact with devices, suggesting a move toward more natural, conversational interfaces.

The keynote is worth a watch. Cristiano hammers home the point that “The biggest coming change is how we will interact with devices.“ He starts digging in around the 14-minute mark, and he highlights several use cases that he expects to migrate from app to AI interaction including travel research and home improvement.

Cristiano’s framing rings true to me. I’m not sure if it’ll arrive in force in the next year or three years, but I’m certain it’s coming.

Playing devil’s advocate, a close friend challenged that assertion. He observed that the migration to voice interfaces was going to be game changing when Alexa emerged over 10 years ago, and that voice interfaces still have not lived up to early hype. It’s a fair point, but I tend to believe that the Alexa phase of voice represents the Palm Pilot era before the iPhone, rather than the first instance of a perpetual shortfall from a technology (e.g. 3D glasses.)

Beyond Qualcomm’s assertions, we’re seeing other evidence of this evolution. Elsewhere in mobile, Deutsche Telekom just announced a new AI phone in partnership with Perplexity at Mobile World Congress this year. Related to living room interfaces, Google announced that Gemini AI will be integrated into Google TV devices with far-field microphones to enable ambient conversational interface. It’s coming.

The AI Interface Revolution

With ChatGPT, Anthropic, and others, we're seeing powerful AI orchestration of underlying agents, commonly accessed and enabled by conversational interfaces. As that matures, we're going to see an increasing shift from discrete apps to APIs for the services that consumers and enterprises rely on. Some envision this evolution as unlocking “ambient computing” where interaction no longer occurs through apps.

To reiterate the point above, I don’t see apps “going away”, but I am confident that reliance on them will diminish over time. Understanding that is important because the AI evolution introduces new architectures, where content and data publishers need to shift focus to exposing their value proposition (to AI platforms) via schemas and APIs, rather than on building standalone UIs.

AI Engine Optimization: A New Frontier

We’re already seeing that happen broadly. There are many, many emerging AI technology providers that offer to ingest content (websites, repositories, PDFs, images, videos, music) and provide summarized versions of that content (often in JSON format) to make it more useful to AI engines.

Just as Search Engine Optimization (SEO) became crucial for digital visibility of websites, "AI Engine Optimization" will become essential in the emerging AI era. Content creators, particularly in industries like entertainment, must prepare their operations and their libraries for AI discoverability. Karen Pfeifer does a good job underscoring that need in a recent post on Medium.

Conclusion

I dedicate considerable time thinking about how technology trends like these apply to Hollywood. For what it's worth, I don’t believe Hollywood has given deep enough thought to how they need to evolve their libraries of content (and the content packages they deliver to distribution partners) to maximize discoverability and utility of films and episodic content in an AI world.

One possible solution to that is what I refer to as Media Resonance Indexing, and it's fodder for a future post.

Before closing, though, I'll note that media indexing for AIEO will be more powerful and useful if it’s a standard. I see reasons that’ll be tough to come by on an industry-wide basis. (That said, MovieLabs has done a lot of great work driving such standardization, and it might play a role here.) Short of that, I expect several holders of significant libraries to work with the major tech companies (who own TV operating systems) to serve as vanguards and seed an approach that evolves into a defacto standard.

Footnote

After I completed this article, I happened upon two relevant items that I’ll share here:

Over the weekend, Joanna Stern published I Quit Google for ChatGPT and Other AI Search – and I’m Not Going Back. It’s a great summary that underscores an underlying theme here.
Amazon Labs just announced Nova Act, an AI agent system that can operate websites on a user’s behalf, making those websites more of an API of sorts for AI, rather than user interface. It is expected to be connected to the next Alexa+ upgrade, bringing together voice interfaces and advanced, agentic AI capabilities.