What are the most and least promising categories to invest in, within developer tools and infrastructure?

The current investment landscape for developer tools and infrastructure is being reshaped by artificial intelligence, which is framed as a new, fourth pillar of infrastructure alongside compute, networking, and storage . This has triggered a massive, non-zero-sum market expansion where value is accruing across the entire stack, from chips and cloud providers to models and applications, making fears of commoditization premature [2, 7]. In this formative stage, the lines between infrastructure and application companies are blurring, with firms like OpenAI succeeding with hybrid, vertically-integrated models that serve both developers and end-users [14, 23]. This dynamic challenges traditional investment theses that separate horizontal platforms from vertical applications, suggesting that defensibility is evolving beyond technical complexity to include distribution, the ability to capture developer attention, and superior integrated experiences [2, 22]. Some venture firms have even bifurcated their enterprise funds to address the distinct diligence processes required for technical versus business software buyers [9, 19].

The most promising investment categories are those directly enabling or built upon this AI-driven shift. AI coding assistants and code generation tools represent a particularly fertile ground, having grown from minimal revenue to a **multi-billion dollar ARR business** in just over two years . Experts see this as the primary domain where long-horizon AI agents are achieving success , with some predicting that the percentage of machine-generated code accepted by companies will eventually reach 100% . This trend is not only boosting developer productivity but also unlocking a vast new market of non-technical domain experts who can now create software, potentially expanding the total software market far beyond its current state [6, 15]. A new frontier is also emerging in "AI-native" infrastructure designed specifically for AI agents, which are becoming a distinct class of economic actors requiring machine-readable documentation and dedicated services . Tools enabling "vibe coding," such as Replit and V0, are also identified as key companies to watch .

Adjacent infrastructure categories are also poised for significant growth. As AI-generated code becomes ubiquitous, tools for automatic security analysis and languages with strong safety guarantees will become more valuable . Similarly, infrastructure-adjacent SaaS companies providing the essential "plumbing" for the AI application layer are finding renewed relevance and growth opportunities . The physical components of the AI buildout, particularly the energy and industrial supply chain for data centers, are now primary bottlenecks and thus a critical investment area . The go-to-market strategy for these tools is crucial, with a bottoms-up, product-led growth model proving highly effective as it allows individual engineers to adopt tools quickly, bypassing slower, top-down enterprise sales cycles that often stall on security and privacy concerns [3, 8].

Conversely, the least promising category appears to be traditional application-layer SaaS, which faces an existential threat from AI's disruptive capabilities . A prevailing theory suggests that companies will increasingly use AI tools to generate niche software applications internally on-demand, reducing the need to purchase standardized products from third-party vendors . This creates a market bifurcation where infrastructure-focused SaaS thrives while application-focused SaaS is disrupted . A key tension for investors is monitoring whether foundation model providers will remain infrastructure partners or move up the stack to compete directly with application companies [17, 29]. Finally, while the enterprise AI opportunity is massive, its adoption will likely be a **slow, multi-year process** due to significant hurdles in redesigning workflows, ensuring regulatory compliance, and managing security, a reality that contrasts with Silicon Valley's expectations of rapid, frictionless adoption .