AI Drives Software Creation Demand, Not Its Demise

The "Software is Dead" Myth: Why Engineering and Product Teams Are the Last Bastions of Growth in the AI Era

The conversation with Mike Cannon-Brookes, co-founder of Atlassian, on The Twenty Minute VC podcast reveals a stark reality: while many sectors face existential threats from AI, the core of software development--engineering and product--is not only surviving but thriving. This discussion unpacks the non-obvious implications of AI adoption, highlighting how the relentless pursuit of efficiency in non-engineering functions creates a compounding demand for sophisticated software solutions. Those who understand this dynamic gain a significant advantage by focusing on building the tools that enable this new era of creation, rather than those that merely automate existing, shrinking processes. This analysis is crucial for founders, investors, and leaders navigating the seismic shifts in the tech landscape, offering a strategic lens to identify enduring value amidst the current AI-driven frenzy.

The Unseen Demand: How AI's Efficiency Drives Software Creation

The prevailing narrative often paints a picture of AI as a disruptive force poised to replace vast swathes of the software industry. However, insights from Mike Cannon-Brookes, co-founder of Atlassian, on The Twenty Minute VC podcast suggest a more nuanced reality. While many roles and software categories face contraction, the core engine of innovation--product and engineering teams--is experiencing a renaissance, driven by AI's ability to accelerate creation. This paradox creates a unique opportunity for companies that can tap into this amplified demand for sophisticated tooling.

The conversation pivots around the sheer scale of predicted AI revenue, with figures like Anthropic's $149 billion ARR in 2029 being a focal point. While impressive, the immediate question arises: where does this money come from? If it's a zero-sum game, it implies a cannibalization of existing software budgets. However, Cannon-Brookes argues that the bet is on TAM expansion. The critical insight here is understanding how AI drives this expansion. It's not just about automating existing tasks; it's about enabling the creation of more software, more complex products, and more services.

"The idea that software as a category is dead is ludicrous to me. It's tough out there, it's going to be hard. Welcome to the technology industry. I just think we have to give up on TAM. I think we just have to let the revenue show us the path to TAM. I think every category that I know of outside of engineering and product is at existential risk of shrinking seats."

This quote encapsulates the core argument: AI, particularly through tools like LLMs that assist in coding, is fundamentally changing the velocity of software development. Engineering and product teams, far from being diminished, are becoming more productive, leading to an explosion in the creation of new software. This, in turn, creates a direct demand for the very tools that Atlassian provides--platforms for managing issues, tracking progress, and collaborating on complex projects. The immediate benefit is amplified output from these core teams, which translates into increased adoption of their platforms.

The downstream consequence of this amplified creation is a ripple effect across the business landscape. While non-engineering functions like customer support, sales, and HR might see seat reductions as AI automates their tasks, the output of engineering and product teams constantly generates new needs and new markets. This creates a dynamic where the "obvious" solutions--those that automate existing, shrinking functions--are vulnerable, while those that enable creation and manage complexity are poised for growth.

Consider the stark contrast drawn between Atlassian and traditional support software like Zendesk. While Zendesk might be seeing pressure as AI automates human support roles, Atlassian benefits because the AI-driven acceleration of software development necessitates more sophisticated management of that development process. More code means more issues, more tickets, more collaboration needs--all of which are bread and butter for Atlassian's suite. This isn't just about incremental improvement; it's about a fundamental shift in the productivity curve for software creation.

The implication for competitive advantage is clear: companies that focus on building the foundational tools for this new era of accelerated software creation, rather than those that optimize for the diminishing returns of automating existing, non-creative roles, will build moats. This requires a long-term perspective, understanding that the immediate discomfort of AI-driven automation in some sectors is precisely what fuels the demand for advanced software solutions in others. The "aha moment" for customers, as Cannon-Brookes implies, is shifting from simply automating a task to enabling entirely new forms of creation and problem-solving.

The "Death of Software" Delusion: Why SaaS Endures

The narrative surrounding the "death of software" is a persistent, yet, according to the discussion, fundamentally flawed, interpretation of AI's impact. While acknowledging that not every SaaS company will survive the next decade, the core argument is that software as a category is not only alive but is being fundamentally reshaped and, in many areas, revitalized by AI. The confusion arises from a failure to distinguish between different types of software and their roles within the broader business ecosystem.

Cannon-Brookes points out that the technology industry has always seen churn. Competitors from years past have disappeared, been acquired, or faded into obsolescence. This is the natural lifecycle of capitalism, exacerbated by technological shifts. However, the idea that AI will render all software obsolete is a gross oversimplification. Instead, AI is becoming a powerful new layer within software, enhancing its capabilities and driving new forms of value.

"The idea that software is a category is dead is ludicrous to me. I'm like, where's the second? It's very efficient for businesses to buy pre-canned solutions of technology. They don't write everything with assembly and they probably still won't."

This highlights the enduring need for packaged solutions that solve business problems efficiently. While AI might change how software is built and what it can do, the fundamental value proposition of SaaS--providing specialized, scalable solutions--remains. The key distinction emerges between software that automates existing, human-centric workflows that are being reduced (like traditional customer support or administrative tasks) and software that enhances human creation and problem-solving (like development tools, advanced analytics, or specialized AI platforms).

The discussion around Harvey, the legal tech AI company, exemplifies this. Despite the potential for AI to automate legal tasks, Harvey's rapid growth and substantial valuation suggest that AI is not simply replacing legal software but is augmenting it, creating new efficiencies and capabilities that expand the market. The argument is that while some software categories will shrink, the overall demand for technology that drives productivity and innovation will continue to grow, albeit with a significant re-allocation of capital and focus.

The concept of RPO (Remaining Performance Obligation) is brought up as a proxy for long-term customer commitment, with accelerated RPO growth signaling confidence in future value. This suggests that customers are not merely adopting AI as a short-term fix but are making multi-year bets on technology that fundamentally enhances their operations. The "death of software" narrative fails to account for this deeper integration and the ongoing evolution of software's role in business.

The Enduring Value of Product and Engineering: Islands of Stability

The podcast conversation consistently returns to a core insight: while many business functions may shrink due to AI-driven efficiencies, product and engineering teams are becoming even more critical. This is because AI, rather than replacing these teams, is augmenting their capabilities, allowing them to build more, build faster, and tackle more complex problems. This creates a unique "island of stability" for companies that serve these functions.

Cannon-Brookes emphasizes that Atlassian's own growth is partly fueled by the fact that product and engineering teams are beneficiaries of AI spend. As these teams become more productive, they generate more output, which in turn requires more sophisticated tools for management, collaboration, and tracking. This isn't a zero-sum game where AI takes from one area to give to another; it's a paradigm shift where AI amplifies the creative potential of engineering and product, thereby increasing the demand for their supporting infrastructure.

"We're product and engineering and software teams is a significant part of our business, 40%. It's a big part of our business. Service in general is a big part of our business. So your argument about Zendesk service collection is a very large business. It's our largest at-scale business, growing very fast, faster than the overall company."

This highlights a crucial distinction. While traditional customer support roles might be susceptible to automation, the "service collection" business within Atlassian, which likely encompasses IT service management and employee service management, is growing rapidly. This suggests that even in service-oriented functions, the focus is shifting from simply answering questions to managing complex internal workflows and processes, which still requires sophisticated software solutions.

The implication is that companies that can effectively harness AI to empower their product and engineering teams will see a compounding benefit. They will not only be able to build more innovative products but will also require more robust tools to manage that increased output. This creates a durable competitive advantage for platforms that are deeply integrated into the software development lifecycle. The "obvious" solutions that automate shrinking functions will face headwinds, while those that enable the engine of creation will find themselves in a position of increasing strategic importance.

Actionable Takeaways

• Focus on Enabling Creation, Not Just Automation: Prioritize building tools and platforms that empower product and engineering teams to create more and faster. This is where the sustained demand will lie.
• Develop AI-Native Solutions: Integrate AI capabilities directly into your product offerings to enhance functionality and create new value propositions, rather than just bolting on superficial AI features.
• Understand the Shifting TAM: Recognize that AI is expanding the total addressable market for software, particularly in areas that drive innovation and productivity in engineering and product development.
• Build for Complexity: As AI accelerates software creation, the need for robust tools to manage complexity, collaboration, and workflows will only increase.
• Invest in Long-Term R&D: Allocate significant resources to fundamental research and development in AI, even if immediate profitability is not apparent. This is crucial for long-term relevance.
• Embrace the "Above the Fold" Strategy: Identify and invest in software categories that are direct beneficiaries of AI-driven productivity gains, rather than those that are primarily automating existing, shrinking roles.
• Champion Efficient Capital Allocation: For public companies, balance the need for long-term AI investment with short-term financial discipline, clearly communicating the strategic rationale for R&D spend to stakeholders.

Key Quotes:

"The idea that software as a category is dead is ludicrous to me. It's tough out there, it's going to be hard. Welcome to the technology industry. I just think we have to give up on TAM. I think we just have to let the revenue show us the path to TAM. I think every category that I know of outside of engineering and product is at existential risk of shrinking seats."

-- Mike Cannon-Brookes

"The idea that software is a category is dead is ludicrous to me. I'm like, where's the second? It's very efficient for businesses to buy pre-canned solutions of technology. They don't write everything with assembly and they probably still won't."

-- Mike Cannon-Brookes

"We're product and engineering and software teams is a significant part of our business, 40%. It's a big part of our business. Service in general is a big part of our business. So your argument about Zendesk service collection is a very large business. It's our largest at-scale business, growing very fast, faster than the overall company."

-- Mike Cannon-Brookes