The SaaS Landscape Is Shifting
For more than a decade, Software-as-a-Service (SaaS) dominated the technology industry. Its scalable growth, low marginal costs, and user-centric models made it a darling of investors and enterprises alike. However, the narrative is rapidly changing. Artificial Intelligence (AI) is no longer just a complementary tool for software—it’s becoming a full-fledged substitute. This evolution is forcing markets and companies to reassess the future of SaaS.
Recent developments have highlighted this turning point. When Anthropic introduced a suite of AI plugins for its Claude model—capable of performing legal research, sales, marketing, finance, data analysis, customer service, and scientific inquiry—the market reacted fiercely. Within a day, approximately $285 billion in market capitalization vanished from software and service companies. Major players like Thomson Reuters and LegalZoom saw their stock prices plummet, while the broader Nasdaq felt the ripple effect.
The Rise of Foundation Models
The emergence of advanced AI systems like Claude and GPT is altering the software stack. These models are evolving into universal interfaces capable of executing tasks that previously required multiple specialized tools. As AI capabilities expand, traditional software categories face obsolescence.
Unlike the 2016 SaaS correction—where companies competed with each other—2026 presents a different challenge: SaaS firms are now competing with general-purpose intelligence. This paradigm shift is prompting enterprises to reconsider their technology investments. The question is no longer “Which software should we buy?” but rather “Can AI do this already?”
Enterprise Behavior Reflects the Change
Organizations are already adapting. The average large enterprise today uses fewer SaaS tools than it did just a few years ago. Budgets are tightening, and procurement discussions increasingly focus on whether AI can replace traditional applications. This doesn’t signal the end of software, but it does suggest a wave of consolidation and reprioritization.
Public SaaS growth has steadily declined since late 2021. While some attribute this to macroeconomic conditions, the deeper truth lies in the changing nature of enterprise technology consumption. The old model—where software was sold per seat or per department—is increasingly being replaced by AI-driven solutions that offer broader functionality at lower costs.
Capital is Chasing Scarcity
As software margins compress, capital is flowing toward areas where scarcity drives value. In 2025, more than 90% of the $111 billion raised by Silicon Valley scaleups went to AI-related ventures. But even within AI, focus is shifting. While earlier investments targeted chatbots, copilots, and workflow automation, today’s largest checks are directed at physical AI applications.
These include manufacturing, robotics, defense, and energy systems—areas where AI enhances real-world operations. Investors are betting big on these sectors not because they’re trendy, but because they’re essential and constrained by physical realities like labor shortages, supply chain fragility, and energy limitations.
AI Meets Industry: The Rise of Physical Intelligence
Physical AI is gaining traction. Figure, a startup developing humanoid robots for industrial use, recently raised $1 billion in Series C funding. Tesla anticipates producing thousands of such robots by 2025, scaling to half a million units by 2027. Divergent, a company applying AI to industrial fabrication, raised $290 million to automate complex manufacturing processes.
These investments reflect a broader industrial resurgence powered by AI. Predictive maintenance, quality control, and logistics optimization are no longer nice-to-have features—they are mission-critical. As Elon Musk recently pointed out, “chip production will probably outpace the ability to turn chips on,” highlighting the urgent need for energy infrastructure to support AI growth.
The Energy Constraint
Energy is becoming the bottleneck in AI expansion. Despite massive investments, data centers are being delayed or canceled due to supply issues. In the past month alone, over 25 data center projects were postponed, double the number from six months ago. Electricity demand is rising for the first time in 15 years, forcing governments and private investors to collaborate on national infrastructure projects.
Major tech firms are now building AI infrastructure in partnership with governments, particularly in areas like semiconductor manufacturing and energy independence. Venture capital is increasingly aligning with public institutions to tackle these “hard problems”—projects too capital-intensive and long-term for private markets to handle alone.
Atoms Over Bits: A New Investment Thesis
Marc Andreessen’s observation that “the next phase of value creation isn’t in bits alone, it’s in bits and atoms” encapsulates the current investment mindset. Software may be abundant, but physical systems retain pricing power because they’re constrained by reality.
This shift is evident in the acquisition of legacy manufacturing firms. Across the U.S., around 125,000 boomer-owned industrial companies—many with no succession plan—are coming to market. These aren’t tech startups; they’re precision machine shops and specialty manufacturers essential to defense and industrial supply chains. Buying them offers unique, defensible positions in increasingly strategic industries.
In a world where AI is approaching zero marginal cost, value accrues to whatever can’t be easily replicated. That means physical assets, energy systems, and real-world production capabilities. The future is not just about intelligent software—it’s about embedding that intelligence into the physical world.
This article is inspired by content from Original Source. It has been rephrased for originality. Images are credited to the original source.