In 1993, Marc Andreessen, then a young visionary with ambition and a full head of hair, was collaborating at the National Center for Supercomputing Applications (NCSA) at the University of Illinois. Along with a colleague, he developed Mosaic, the browser that revolutionized the way to explore the emerging World Wide Web. Until that point, browsers were text-based and required keyboard commands; Mosaic introduced graphical navigation, allowing for the integration of images and multimedia content. Shortly after, Andreessen met Jim Clark, founder of Silicon Graphics, who encouraged him to start a business. Together, they founded a startup that would give rise to Netscape, one of the most iconic browsers in history. This success made Andreessen a billionaire, and starting in 2005, his focus shifted: he founded the venture capital firm Andreessen Horowitz, investing in new tech companies and coining famous phrases like “Software is eating the world” in 2011. His argument was clear: software companies dominated global growth, a prediction that has proven true. However, the rise of artificial intelligence has changed the landscape. The launch of ChatGPT in November 2022 marked a turning point, although tools like GitHub Copilot in 2021 had already anticipated the potential of generative AI. These solutions allow machines to program, democratizing development and enabling more people to create their own applications, although human oversight is still required. The rise of “vibe coding,” a term that Andrej Karpathy proposed to update to “agentic engineering,” has sparked a frenzy for custom software development. Open-source tools and tailored micro-applications are proliferating, while generative AI models and intelligent agents impact both users and large companies. The SaaS (Software as a Service) model, which transformed applications like Photoshop and Office into subscription-based cloud services, now faces a threat. Last week, major software companies collectively lost $300 billion in market value, with declines of 15-20% in companies like MongoDB, Salesforce, Shopify, and Atlassian. This phenomenon, dubbed “SaaSpocalypse,” reflects market anxiety over AI disruption. These declines are not only due to the advancement of AI but also to customer dissatisfaction with the pricing policies and rigid contracts of SaaS platforms. Examples like Salesforce, with a 35% increase in two years, or Broadcom, with a 1,500% rise in VMware licenses, have pushed customers to seek alternatives, many of which are AI-based, offering efficiency and greater autonomy. The stock market correction can also be seen as a reaction following the 2021 bubble driven by the pandemic. Now, some customers are opting for custom solutions thanks to AI, although this entails risks: implementing systems generated solely by AI requires rigorous human review to ensure reliability. Is AI “devouring” software? Rather, it is transforming it. Steven Sinofsky, former Microsoft executive, recalls that major technological transitions are often perceived as the end of an era, but in reality, they catalyze expansion. The PC did not eliminate the mainframe, nor did e-commerce end physical stores; both evolved and grew. With AI, the demand for software will only increase, as many processes remain to be optimized. The role of the human programmer will not disappear either. AI automates repetitive tasks, but the value will lie in oversight and system architecture. Just as Excel transformed accountants into analysts, AI is turning developers into solution architects. The so-called “SaaSpocalypse” is, in fact, a metamorphosis. Software is shifting from a static product to a dynamic and autonomous service. The challenge will be to deliver results and autonomy, not just lines of code. However, significant challenges arise: the technical debt of relying on AI-generated code, data sovereignty, and security. As Linus Torvalds warned, AI is a powerful tool, but maintenance and deep understanding will remain essential. Additionally, intellectual property and data privacy, along with energy needs and developer training, will be key issues in the future of software.