Is AI destroying the software industry or reshaping it? Let's take a look at some of my analysis below.

1. There are two extreme views on AI in the market right now. The first view holds that as AI becomes increasingly powerful, many jobs will be automated, companies won't need as many employees, and demand for software subscriptions will decline. This leads some to worry that AI will devalue traditional SaaS software.

2. AI will indeed eliminate some traditional software. In the past, many software programs handled information organization, process management, and simple automation. However, with the advent of AI, many tasks that previously required multiple software programs might be solved by a single AI Agent. Therefore, software companies without a moat will face difficulties in the future.

3. The real beneficiaries are AI-native software companies. The future isn't about software disappearing, but about software transitioning from the tool era to the Agent era. In the past, humans operated software; in the future, AI will help humans operate software. Therefore, companies that can deeply integrate AI first may usher in a new growth cycle.

4. The collective rise in software stocks is not accidental. I believe the recent broad-based strength in software stocks—with enterprise software, cloud services, and AI applications all rising—actually represents the market beginning to realize that the beneficiaries of AI are no longer just NVDA and chip companies. Capital is spreading from the computing power layer and chip layer to the application layer and software layer. This is different from last year when capital only chased semiconductors.

5. AI Agent might be the core logic of the next phase. We are gradually moving from the Prompt Engineering era into the Context Engineering and AI Agent intelligent proxy era. Simply put, in the past, AI only answered questions. In the future, AI will execute tasks, call tools, and automatically complete workflows. This will fundamentally change the business models of the software industry. I am personally more optimistic about these areas: enterprise-grade AI software, AI Agent platforms, cloud computing, data management, and enterprise automation, rather than software companies solely reliant on traditional subscription models.

AI won't kill the software industry; it is gradually phasing out old software and giving rise to new software. The recent collective strength in software stocks indicates that capital has already started shifting from chips to software. The next wave of explosive growth might not be Nvda, but rather the software companies that first implement AI Agents in enterprise scenarios. Therefore, everyone can position themselves in advance.$Microsoft(MSFT.US)$NVIDIA(NVDA.US)$Micron Tech(MU.US)