The AI sector has started to cannibalize itself.

Last night, all three major indices were falling, and today the Hong Kong stock market trend isn't looking very good either.

If we keep analyzing the situation and oil prices, it's getting boring; everyone's tired of hearing it. Honestly, I don't even know how to highlight the key points anymore because the logic remains the same: tech stocks will no longer benefit universally like a 'big pot meal', and the earnings season will be an unavoidable window. Hot sectors like tech stocks and AI concepts will undergo internal segmentation. Those who can see profit potential in revenue-generating parts like AI infrastructure and AI creating new business models are the ones worth long-term optimism. If the original business model is being eroded without good transition measures, then sorry, no one might buy into it.

So, combining this logic to look at last night's or even the past few days' market performance becomes clearer. Software stocks $IBM(IBM.US) and $ServiceNow(NOW.US) plummeted, dragging down the entire software and services sector. Similarly, the key point still isn't the earnings report itself. Just looking at the data, both companies delivered revenue and profits above expectations, but they didn't provide answers to more fundamental questions, namely: Will the business models of traditional software companies be gradually eaten away by the new wave of AI tools?

Since Anthropic launched a batch of tools in February this year capable of automatically completing tasks like marketing and data analysis, market anxiety about the software industry has been heating up. So, last night's earnings from IBM and ServiceNow were more like reigniting this concern. It's worth noting that the Software ETF has already fallen about 16% this year.

Then, looking at the other side, $Texas Instruments(TXN.US) surged 19.43% last night, marking its largest single-day gain since October 2000. The reason is simple: its Q2 revenue and profit guidance were above expectations, driven by continued growth in demand for data center-related chips. More crucially, this rise wasn't isolated to individual stocks; the entire chip chain was lifted, with the Philadelphia Semiconductor Index hitting new highs that same day. This means the market isn't losing faith in AI; it's increasingly believing that the ones making real money first aren't those telling application stories, but those selling shovels, power management, and computing infrastructure. While the Software ETF is down about 16% year-to-date, the Semiconductor ETF has already risen over 43%. This divergence itself shows that current capital is actively reprioritizing the industry chain with real money.

If Texas Instruments was just intraday confirmation, then the after-hours earnings from $Intel(INTC.US) pushed this narrative a step further. Intel's Q1 revenue of $13.58 billion beat market expectations, and its Q2 revenue guidance of $13.8 to $14.8 billion was also significantly above Wall Street estimates, sending its stock up nearly 19% after-hours. What does this show? It shows that even though Intel hasn't been a core player in the AI narrative for the past few years, as long as it can now prove it's capturing demand from AI data centers and server CPUs, the market is willing to reprice it.

So, if today's article still only writes generic stuff like "Middle East disturbances risk appetite" or "US stocks oscillating at highs," I think it misses the point. The real key point is: The AI theme has moved from a unified valuation story into a stage of reshuffling by industry, segment, and business model.

Therefore, my own understanding is that when looking at tech stocks next, focus on a core essence: Are you selling AI, or are you being replaced by AI? In the former, the first to be acknowledged by the market is still the line of chips, servers, data center power, and infrastructure. In the latter, even if some traditional software companies are still profitable, they may face sustained pressure due to loosening valuation logic. Coupled with oil prices remaining high and market inflation concerns unresolved, it's actually quite difficult for the broader market to continue mindlessly charging upwards. So, the most important thing today isn't fearing a decline, but not treating all tech stocks as one basket. AI hasn't fizzled out; it's just starting to be selective.

The above content represents personal views only and does not constitute investment advice. Give Vivian a follow, and I wish everyone's buys go up and sells go down~