The current tech landscape is not a universally rising tide. While investor enthusiasm buoys AI-native companies, the disruptive threat of large language models is simultaneously depressing valuations and venture capital interest for traditional software companies whose business models are now at risk.

The market is simultaneously devaluing software companies because AI is a viable competitor, while also punishing AI infrastructure companies for their massive capital expenditures with uncertain returns. This contradictory fear creates broad, indiscriminate selling.

IBM's stock plunged 13% after an Anthropic blog post about an *existing* AI capability (COBOL modernization). This indicates investors are finally grappling with the long-term disruptive implications of AI on legacy businesses, reacting to the strategic threat itself rather than waiting for a specific new product launch.

Investors are selling off hyperscalers like Amazon for their massive $200B AI CapEx, fearing pinched profits. Simultaneously, software stocks are being punished for not investing enough in AI. This contradictory reaction highlights extreme market uncertainty about the right AI investment strategy.

Arvind Krishna predicts that the largest AI models will become commodities with low switching costs. This belief underpins IBM's strategy to *not* compete in building frontier models, but rather to partner with providers and focus on smaller, specialized enterprise models where they can build a moat.

IBM's CEO argues the AI bubble is in data center construction. The committed build-out requires an additional $1-2 trillion in new annual revenue to justify the investment—a figure he believes is unrealistic, meaning many infrastructure bets will fail.

IBM CEO Arvind Krishna's strategy rests on the conviction that most enterprises will remain hybrid, avoiding lock-in to one public cloud. This creates a durable market for IBM's management software. The second pillar is focusing on deploying trusted AI in regulated industries, ceding the consumer space to others.

The AI productivity boom is not lifting all tech stocks. Instead, it's negatively impacting traditional software companies. The market is pricing this in, with software ETFs like IGV breaking down technically even before earnings reports reflect the anticipated decline in business.

Incumbent software vendors face a crisis: customers aren't churning, but all new enterprise budget is directed at AI. This traps legacy platforms as stagnant 'systems of record' while AI applications built on top capture all future growth.