Communications Fireside Chat w/ Hedgeye Asset Management Portfolio Manager Sam Rahman (1/22)

Key Concepts

• AI as a New Abstraction Layer: AI-powered personal assistants are poised to become the next fundamental shift in human-technology interaction, potentially diminishing the importance of traditional apps and operating systems.
• Google-Apple Partnership: A landmark strategic alliance integrating Google’s Gemini models into Apple’s foundation models, driven by antitrust context and mutual benefits in AI capabilities and distribution.
• Disruption of Ecosystems: AI agents threaten existing app ecosystems, traditional advertising models (especially performance marketing), and e-commerce, shifting value towards direct agent-to-agent interactions.
• Bifurcated AI Future: The emergence of large, general "world models" (dominated by hyperscalers) alongside specialized models leveraging proprietary data.
• Platform Control & Distribution: The critical importance of owning the customer relationship, controlling distribution (hardware ecosystems), and having robust infrastructure and scale for AI dominance.
• Market Dynamics & Investment: A "winner-take-most" environment with significant risks for companies not engaging with open AI protocols, requiring caution ("falling knife") and a focus on secular trends and execution.

The Google-Apple Partnership and the AI Abstraction Layer

The core discussion revolves around the significance of Google’s partnership with Apple to integrate Gemini models into Apple’s foundation models, effectively white-labeling Google’s AI for Siri and potentially a full-fledged chatbot akin to OpenAI’s ChatGPT. This is viewed as a landmark event, potentially reshaping human-technology interaction. This partnership was heavily influenced by the recent DOJ ruling regarding Google’s potential monopoly and its long-standing search relationship with Apple, with Apple strategically choosing Google due to its superior AI technology.

A central theme is the concept of a new “abstraction layer” in technology. Historically, disruptions like the PC, internet, and streaming video represented these shifts. The speakers believe AI, specifically AI-powered personal assistants, could become the next abstraction layer, potentially diminishing the importance of traditional apps and operating systems. This shift means that controlling the customer relationship and pricing power are key benefits of owning this layer.

Impact on App Ecosystems and Advertising

The integration of AI agents raises concerns about the future of the app ecosystem. If AI agents handle tasks directly, app discovery and engagement could decline, particularly for non-essential apps and ad-supported models. An example of booking a trip illustrates how an AI agent could streamline the process, potentially bypassing traditional apps and websites like Concur. This shift extends to e-commerce, where AI could empower smaller merchants and challenge Amazon’s dominance.

The discussion highlights the shift towards “agent-to-agent protocols,” where AI agents interact directly with each other. This is expected to disrupt traditional advertising models, particularly performance marketing and display ads. While acknowledging the potential for auction dynamics within these protocols (allowing for “paid promotion” of products), concerns are raised about monetization and fairness, drawing parallels to issues experienced with music streaming platforms like Spotify. The future of advertising will be about “real estate” – maintaining relevance and user base. Companies like The Trade Desk and Reddit are mentioned as facing disruption in this evolving ad landscape.

The Bifurcated AI Future

Sam Ramen posits a future with two types of AI models: large, world models (dominated by companies with massive capital like Google) and specialized, smaller models leveraging unique, proprietary data (like JP Morgan’s financial data or Apple’s user data). While Large Language Models (LLMs) may become commoditized, the value will shift to what is built around the models – specifically, AI agents accessible via APIs.

Company-Specific Analysis

• Google: Positioned as significantly better than competitors due to its existing infrastructure, early investment in AI (DeepMind), and recent execution improvements spurred by a “code red” situation and the return of its founders. Its dominance in search, achieved through integration with iOS, illustrates the power of distribution. Google’s lead is substantial, and it is strategically aligned with Apple despite potential antitrust concerns.
• Apple: Its strong hardware ecosystem and loyal user base provide crucial distribution for Google’s AI, while Google provides Apple with the AI capabilities it lacked. Apple’s capex as a percentage of revenue is described as “super low.”
• Amazon: While possessing a strong ecosystem, Amazon faces a discoverability risk if it doesn’t integrate with open AI protocols. Its large ad business, currently reliant on display and sponsored ads, may be challenged by agent-based systems not optimized for those formats.
• Meta: Described as a “higher beta” business model, heavily reliant on external platforms, lacking control over its destiny, and structurally different from platform companies like Google. While Meta boasts impressive 45% operating margins, future investments (like Reality Labs and Meta Compute) will likely be margin dilutive. The success of Meta’s bet on augmented reality glasses is seen as crucial; a significant decline in on-device time (e.g., smartphone usage) would be detrimental, as AI's potential impact on time spent on platforms is considered an existential risk.

Investment Perspectives and Market Dynamics

The speakers reminisce about the exceptional performance of growth stocks in the 2010s, driven by low interest rates, disruption, and new technology, suggesting that understanding these secular trends is crucial for investment success. The current market conditions are described as a “fastly falling knife,” advising caution against premature investment. The “winner-take-most” dynamic observed with the internet is anticipated to recur, leading to a potentially painful market correction as companies vie for dominance in the AI space.

Execution is paramount, with Google’s recent turnaround cited as a prime example. Having robust infrastructure and scale (like Google and AWS/Azure hyperscalers) is critical for success in the AI era. Companies that control their own platforms (like Google and Apple) are better positioned than those reliant on others (like Meta). Missing the next big platform shift is considered the “death nail” for any major technology company over time.

Technical Terms & Concepts

• LLM (Large Language Model): A type of AI model trained on massive datasets of text, capable of generating human-like text.
• Foundation Models: Large AI models that can be adapted for a variety of tasks.
• Abstraction Layer: A simplified interface that hides the complexity of underlying systems.
• Multimodal AI: AI models that can process and generate multiple types of data (e.g., text, images, video).
• Capex (Capital Expenditure): Funds used by a company to acquire, upgrade, and maintain physical assets.
• JSON Manifest: A standardized format for describing data, used to define services offered by a merchant.
• MCP (Model Context Protocol): A new protocol for agent-to-agent communication and interoperability.
• Inference: The process of using a trained AI model to make predictions or generate outputs.
• Agent-to-Agent Protocols: Systems where AI agents communicate and transact directly with each other.
• API (Application Programming Interface): A set of rules and specifications that allow different software applications to communicate with each other.
• Hyperscaler: A company that provides a massive scale of cloud computing resources.
• GCP (Google Cloud Platform): Google’s suite of cloud computing services.
• Operating Margins: A measure of a company’s profitability, calculated as operating income divided by revenue.

Data & Statistics

• Google’s Search Market Share: 80-90% (mentioned as a historical example of dominance).
• Apple’s Capex as a Percentage of Revenue: Described as “super low.”
• The 2010s: Characterized as the “best bull market for growth investors ever.”
• Meta’s Operating Margins: 45% (remarkably high for a company of its size).
• Potential Decline in On-Device Time: A hypothetical 50% decrease in time spent on smartphones over the next 15 years.

Notable Quotes

• “If you're able to go back in time 10, 12 years and really understand what these businesses were going to become over the coming 10 years, I think you would have been long most of these names and that would have made most of your performance.” – Sam Ramen
• “The best bull market for growth investors ever.” – Andrew Freeman (referring to the 2010s)
• “The reason why search became such a dominant presence…was once they got onto the iOS platform that was game over for any other competition because it's all about distribution.” – Sam Ramen
• “Don’t catch a falling knife.” (Investment advice in the current market)
• “It’s like it’s real estate and you know, can you stay relevant in the future?” (On the importance of maintaining user base)
• “Missing the next big platform shift is the death nail for any major technology company over time.” (On the importance of innovation)

Conclusion

The discussion underscores that AI represents a fundamental, transformative shift, creating a new "abstraction layer" that will redefine human-technology interaction and disrupt established ecosystems. The Google-Apple partnership exemplifies the strategic alliances forming in this new era, driven by the critical need for both advanced AI capabilities and robust distribution. While the future of AI is likely bifurcated between large general models and specialized proprietary ones, success hinges on platform control, robust infrastructure, and agile execution. Investors are advised to navigate this "winner-take-most" environment with caution, focusing on companies that can adapt to the evolving landscape of agent-to-agent protocols and maintain relevance in the face of existential threats to traditional business models.