"It's Different This Time!"
From Panic to Perspective: The Real Impact of AI on Enterprise Software
One of the most common expressions of fear that I have encountered in 25 years of investment management is the phrase: “It’s different this time!” It presents a justifiable claim that the present fear is rational—that the worry or panic over some existential threat is valid.
This reaction is understandable. And short-sighted. The fear displayed is myopic—emphasis on MY! “My world is coming to an end!”
Of course, “the world is ending” is hyperbole. What fear-mongers really mean is that the way things—as we knew them—are changing. But the world is always changing. Stability is the illusion; adaptation is the constant.
So when the software sector slips into bear-market territory as artificial intelligence disrupts once-untouchable growth assumptions, some investors flee for higher ground. The software sector as we knew it is finished. “It’s different this time!”
The Meltdown Narrative
The current story predicting the imminent demise of enterprise software—particularly subscription-based SaaS—runs something like this:
Enterprise software flourished under a clean, elegant economic logic: high switching costs, recurring revenue, durable cash flows, enviable margins.
AI systems now replicate large swaths of human work—research, drafting, coding, analysis, coordination—without needing to live inside proprietary applications.
Customers will gradually abandon legacy platforms; pricing power erodes, cash flows weaken, margins compress.
In short: AI is framed as a dissolvent of traditional software moats. The golden age is over. For investors heavily allocated to software, this is not merely unsettling—it is existential.
But this narrative rests on a crucial assumption: that technological substitution automatically implies economic annihilation.
History suggests otherwise.
Collapse or Reallocation?
What if the software sector isn’t dying—but being re-priced around a different center of gravity?
Enterprise software spending today sits near $1.24 trillion globally. According to research from Goldman Sachs, the total addressable market for software could expand 20% to 45% by 2030, driven not by the disappearance of software, but by its evolution toward agentic systems—AI models embedded within workflows, APIs, data layers, and governance frameworks.
This is a picture of value migration rather than shrinkage.
AI does not abolish systems; it intensifies the demand for integration, orchestration, compliance, security, auditability, and domain-specific reliability. Models may be general, but enterprises are not.
The lesson from the early internet era is instructive. In the late 1990s, brick-and-mortar retail was pronounced dead. Middlemen were finished. Geography was irrelevant. Yet the ultimate winners were not disembodied websites, but integrated platforms—logistics, payments, data, supply chains—wrapped around the new technology.
The same pattern is unfolding again.
The “Real World” Friction Everyone Forgets
There is another assumption embedded in the “SaaSpocalypse” thesis: that AI adoption will be swift, seamless, and immediately accretive to productivity.
Evidence suggests the opposite.
Recent work highlighted by Harvard Business Review shows that AI frequently intensifies work rather than eliminating it—creating new layers of oversight, verification, coordination, and exception-handling. Meanwhile, engineers and knowledge workers are already reporting cognitive overload and diminishing returns as AI tools proliferate faster than organizations can absorb them.
A growing phenomenon called “AI fatigue” is resulting in productivity bottlenecks, trust gaps, and workflow fragmentation caused by too many semi-autonomous tools competing for attention rather than cooperating within a coherent system.
This matters because enterprises do not buy capabilities in isolation. They buy outcomes—and outcomes require stability, accountability, and integration. All of which slow adoption, complicate deployment, and favor incumbents who already sit inside the workflow.
AI will be transformative. It will also be messy, delayed, uneven, and politically constrained inside organizations.
What Actually Changes
The most likely outcome is not the extinction of software companies, but a reshuffling of where economic values accrue:
Standalone point solutions face compression.
Undifferentiated SaaS tools lose pricing power.
Integrated platforms—those that own data gravity, workflows, compliance, and distribution—gain leverage.
Software becomes less about screens and more about systems.
Adaptation will result in more normalized margins and valuations. Likely the “growth” narrative will go through maturation. That is what disruption looks like.
Smart investors adapt alongside these changes.
A Familiar Ending
Every technological revolution produces its own version of “It’s different this time.” Railroads. Electricity. The internet. Smartphones. Cloud computing.
Each time, pessimists mistake transition for termination. Change makes it feel like the end of the world, but it often becomes an engine for growth.
AI will change software profoundly. It will punish complacency and reward integration. It will expose weak moats and deepen strong ones. But the idea that enterprises will abandon structured systems in favor of free-floating intelligence misunderstands how organizations function—and how value is actually captured.
So yes, this time is different. Just not in the way many insist it must be.
This publication is for brains, not bets. The Other Side of Obvious shares ideas, stories, and general financial information - not personalized investment, tax, or legal advice. Investing comes with risk (including losing money). Talk to a pro before you act. Please take time to read these important disclosures before you get started.