AI Disruption Reshapes Markets--Value Accrues to Agents

The AI Disruption: Beyond the Hype, Towards Durable Advantage

This conversation reveals a stark reality: the rapid advancement of AI is not just an incremental improvement but a fundamental reshaping of the software landscape, creating hidden consequences for incumbents and opportunities for those who embrace the shift. The non-obvious implication is that conventional wisdom regarding market dynamics and competitive advantage is becoming obsolete at an accelerated pace. Founders, investors, and product leaders who grasp the systemic nature of this disruption will gain a crucial edge by anticipating downstream effects and positioning themselves for long-term resilience, rather than reacting to immediate market tremors. This analysis is essential for anyone seeking to navigate the current technological paradigm and build enduring value.

The Unseen Cost of AI's Advance: Why "Good Enough" Becomes "Too Late"

The rapid proliferation of AI capabilities, particularly in areas like code generation and security auditing, is not merely an evolution of existing tools; it's a fundamental disruption that redefines competitive moats. While many incumbents are focusing on integrating AI as a feature, the deeper implication, as highlighted by Jason Lemkin, is that the "agentic layer" is poised to capture disproportionate value, potentially commoditizing existing software. The market's reaction to Anthropic's security feature, while perhaps overblown in its immediate impact on stock prices, signals a critical shift: AI is rapidly surpassing the capabilities of even mediocre human engineers, making existing security solutions vulnerable if they are priced for perfection.

"When you are priced for perfection, anything less than perfection will be a kick in the nuts."

This pricing for perfection, combined with the accelerating pace of AI development, creates a dangerous feedback loop. Companies that were once considered safe harbors, like CrowdStrike, are now facing scrutiny not because their product is obsolete, but because the market's expectations have outpaced their AI integration. The analysis suggests that the true disruption isn't about AI replacing an entire industry overnight, but about AI's ability to incrementally erode the value proposition of incumbents by offering "good enough" solutions at a fraction of the cost or complexity. Rory O'Driscoll points out that AI capabilities, like code scanning, have existed in models like Claude for months, yet the market reacts as if it's entirely new. This highlights a systemic blindness to the cumulative impact of AI's progress. The danger lies not in a single AI breakthrough, but in the continuous, compounding improvement that leaves those who are slow to adapt with shrinking market share and relevance.

The Agentic Tide: How Value Accrues to the Orchestrators

The conversation repeatedly returns to the concept of "agents" and the "agentic layer," a critical insight into how value will be captured in the AI era. Jason Lemkin argues that more and more value will accrue to whoever builds the dominant agentic layer, a stark contrast to the current model where value is concentrated within specific software applications. This is particularly relevant when considering consumer-facing platforms like DoorDash. While the CTO of DoorDash, Andy Fang, believes agents will be transformative, Lemkin pushes back, suggesting that the emotional and nuanced decision-making involved in ordering food is not something consumers will readily delegate. However, the underlying threat remains: an agent that can optimize for price, delivery time, and even personal preference by analyzing past behavior and external trends could indeed disrupt the current ecosystem.

"Almost everyone that's public, your surface area is shrinking because of Claude and AI, and the question is how much?"

This shrinking surface area for incumbents is a direct consequence of agents becoming the primary interface for many tasks. The argument isn't that DoorDash will disappear, but that its value proposition could be significantly diminished if an agent becomes the decision-maker for consumers. This dynamic extends to design tools like Figma. While Figma's recent earnings show resilience, the underlying concern is that AI models, like Claude Code, will eventually be able to generate designs that are "good enough," if not superior, to human designers. The "Fortnite circle shrinking" analogy perfectly captures this existential threat: as AI capabilities expand, the territory for existing tools and platforms will inevitably contract. The critical question for businesses is not if they will be disrupted, but how they will adapt to a world where agents orchestrate more and more of their customer interactions and internal processes.

The Delayed Payoff: Building Defensibility in an Age of Acceleration

The discussion around OpenAI's ambitious spending plans and Anthropic's market impact underscores a crucial aspect of AI-driven disruption: the immense capital requirements and the long-term nature of building defensible positions. OpenAI's strategy of investing heavily in hardware, ads, and new ventures, while seemingly unfocused to some, reflects a belief that controlling multiple facets of the AI ecosystem is necessary to maintain leadership. Rory O'Driscoll notes that OpenAI's previous lead has narrowed, suggesting that even aggressive expansion can lead to market share erosion if core areas are neglected. This highlights the tension between breadth and depth in AI strategy.

The insight here is that true competitive advantage in the AI era will likely come from building capabilities that have a delayed payoff, requiring significant upfront investment and patience. For instance, while many B2B software companies are struggling with dated products, those that can successfully integrate AI to create genuinely novel workflows or deeply personalized user experiences will build a moat that is difficult for AI-native disruptors to replicate quickly. Jason Lemkin’s preference for investing in momentum plays, while seemingly counterintuitive given the rapid pace of change, reflects a pragmatic approach to navigating uncertainty. The argument is that in a market where the future is so fluid, betting on companies that are already demonstrating traction and growth--even if their valuations seem high--is a more reliable strategy than trying to find value in beaten-down stocks. This is where the "competitive advantage from difficulty" comes into play. Companies that undertake the hard work of building truly integrated, AI-powered solutions, even if it means a longer time to market or higher upfront costs, will ultimately create more durable moats than those chasing immediate feature parity.

Key Action Items:

• Embrace the Agentic Layer: Identify how agents can become the primary interface for your customers or internal workflows. Prioritize building or integrating with dominant agentic platforms. (Immediate to 6 months)
• Rethink "Good Enough": Continuously assess if your current product offerings will be "good enough" in 12-18 months given AI's accelerating capabilities. Invest in AI-native features that go beyond incremental improvements. (Ongoing, with quarterly reviews)
• Focus on Delayed Payoffs: Prioritize investments in AI capabilities that require significant upfront effort but promise long-term defensibility, rather than quick feature integrations. This might involve building proprietary data moats or complex, integrated workflows. (12-18 months for initial payoff)
• Strategic Partnerships for AI Integration: Actively explore partnerships with AI model providers and agent developers to ensure your products remain relevant and integrated into the evolving AI ecosystem. (Immediate)
• Scenario Planning for Incumbents: For established companies, conduct rigorous scenario planning to understand how AI could erode your core value proposition within 18-24 months. Develop proactive strategies to counter these threats. (Next quarter)
• Invest in Talent for AI Adaptation: Focus on hiring and upskilling talent capable of building, deploying, and managing AI-driven agents and complex AI workflows. This is a longer-term investment in organizational resilience. (12-24 months for significant impact)
• Monitor Market Momentum Selectively: While not abandoning value investing, pay close attention to market momentum in specific AI-driven sectors. Consider strategic bets on companies demonstrating clear traction and growth, understanding the inherent risks. (Ongoing)