AI spending is rising as software and advertising revenues face disruption

The article argues that AI is disrupting the business models of the companies investing heavily in it. AI-assisted cloning could produce functional alternatives to products such as Microsoft Office and Adobe Creative Suite, potentially weakening software firms’ ability to rely on proprietary code and customer lock-in.

The cloned applications are described as incomplete, buggy, and released as clean-room reimplementations, raising difficult questions about intellectual property and the companies’ own use of AI trained on others’ work. The article also highlights a possible advertising “doom loop”, competition from efficient Chinese models, and doubts about returns on AI infrastructure spending. It cites a forecast of $11 trillion in total AI capital expenditure by 2030 and describes how Nvidia and other large technology companies support cash-hungry neocloud firms by guaranteeing demand for computing capacity.