Jensen Huang Pushes Back on 'Saaspocalypse' Fears at Computex 2025
At Computex, one of the world's most prominent technology conferences held annually in Taipei, Taiwan, Nvidia CEO Jensen Huang delivered a message that many software entrepreneurs and SaaS executives desperately needed to hear. Rather than sounding the alarm on artificial intelligence disrupting and dismantling the traditional software industry, Huang made a bold and optimistic case: now is actually an "incredible time" to be a software company. His remarks arrive at a moment of significant anxiety in the software sector, where fears of a so-called "Saaspocalypse" have been weighing heavily on investor sentiment and company valuations alike.
What Is the 'Saaspocalypse' and Why Are Software Companies Worried?
The term "Saaspocalypse" has been circulating in tech and financial circles with growing urgency. It refers to the fear that the rapid advancement of artificial intelligence — particularly large language models and autonomous AI systems — could render traditional Software as a Service (SaaS) platforms obsolete. The logic goes something like this: if AI can perform complex tasks autonomously, why would businesses continue paying for specialized software subscriptions when a single AI agent could replace dozens of point solutions?
This anxiety has not been purely theoretical. Major SaaS companies including Salesforce and Workday have experienced notable stock declines as markets grapple with the existential question of whether AI is a tailwind or a headwind for the software industry. Investors have been asking hard questions about the long-term viability of subscription-based software businesses in a world where AI agents are becoming increasingly capable.
Against this backdrop, Huang's remarks at Computex carry significant weight. As the CEO of the company whose chips power virtually every major AI system in the world, Jensen Huang has perhaps the clearest view of where artificial intelligence is actually headed — and what it means for software.
Jensen Huang's Vision: Agentic AI as a Software Multiplier
"A lot of people have said, 'Jensen, AI is coming. Agentic AI is coming. Therefore, all of the software companies are going to go out of business.' I said it's exactly the opposite," Huang told the audience at the start of his keynote address on Monday.
His counterintuitive argument centers on the concept of agentic AI — AI systems designed to complete complex, multi-step tasks with minimal human intervention. Unlike traditional AI tools that respond to individual prompts, agentic AI systems operate more autonomously, executing workflows, making decisions, and interacting with external tools and services to accomplish goals.
Huang's insight is that this shift doesn't eliminate the need for software — it dramatically increases it. Because AI agents need to interact with the world, they require tools. And more tools means more software. "Because there will be many agents doing work, they will be using more tools than ever," Huang explained.
This framing redefines the software company's role in an AI-first world. Rather than being replaced by AI, software platforms become the infrastructure that AI agents operate within and depend upon. The key, according to Huang, is adaptation: "This is actually an incredible time to be a software company, but the software has to be presented to the agent in a way that the agent can use it."
What Does 'Agent-Ready' Software Actually Mean?
Huang's caveat — that software must be presented in a way that agents can use it — is arguably the most important part of his message for software executives and developers. It signals that the companies that will thrive in the agentic AI era are those that proactively redesign their products to be consumable by AI systems, not just human users.
In practical terms, this means several things for software companies looking to future-proof their offerings:
- API-first architecture: Software must expose clean, well-documented APIs that AI agents can call programmatically to retrieve data, trigger actions, and complete workflows without requiring a human to navigate a graphical interface.
- Structured data outputs: Agents need information in formats they can reliably interpret and act upon. Software that produces structured, machine-readable outputs will be far more valuable in an agentic ecosystem than tools that only generate human-readable reports or dashboards.
- Integration with AI orchestration platforms: As agentic frameworks like LangChain, AutoGPT, and Nvidia's own AI tooling ecosystems grow in adoption, software companies that build native integrations will position themselves as essential components of AI-driven workflows.
- Granular permissions and audit trails: AI agents operating at scale will require software platforms to offer fine-grained access controls and detailed logging, enabling enterprises to maintain oversight and compliance as autonomous systems handle more sensitive tasks.
Why the Computex Stage Matters for This Message
The choice of Computex as the venue for this reassurance was itself meaningful. Computex sits at the intersection of hardware innovation and software application, drawing developers, enterprise technology buyers, chip manufacturers, and AI researchers from around the globe. For Huang to open his keynote by directly addressing SaaS fears signals that this conversation is no longer confined to Wall Street analysts — it is now a defining strategic question for the entire technology industry.
Nvidia, of course, has a vested interest in painting an optimistic picture for the software industry. More software activity means more compute demand, which means more demand for the GPUs that Nvidia designs and sells. But that commercial alignment doesn't make Huang's underlying thesis any less compelling. The history of technology is filled with examples of new paradigms expanding the market for adjacent technologies rather than destroying them — the internet didn't kill software; it created an entirely new category of it.
The Bigger Picture: AI as Expansion, Not Replacement
Huang's broader message aligns with a perspective shared by a growing number of enterprise technology leaders: that AI, when properly integrated, acts as an amplifier for human productivity and organizational capability rather than a simple substitution engine. The companies most at risk in the agentic AI transition are not those that build software, but those that build software which refuses to evolve.
For SaaS companies watching their valuations fluctuate with every new AI announcement, the path forward is not to fear the agent — it is to become indispensable to it. The software companies that will define the next decade will be those that redesign their products around the assumption that their primary user may increasingly be an AI, not a human.
Jensen Huang's "incredible time" declaration is ultimately a challenge as much as it is a reassurance. The opportunity is real and substantial. But seizing it requires a fundamental rethinking of what software is for, who it serves, and how it must evolve to remain relevant in a world where intelligence itself is becoming programmable.