AI has finally become expensive enough to make Wall Street nervous

During earnings season, Google surprised investors by raising its spending estimate to as much as $205 billion, up from a previous projection of up to $190 billion. This increase indicates that Google cannot accurately forecast its costs, and it is spending more than it earns. Google also faces competitive pressure from Chinese AI tools and pricing pressure. These concerns are not limited to Google; Meta, Amazon, and Microsoft are also expected to report higher-than-expected spending on data centers. Investor nervousness is further fueled by a drop in SpaceX shares, Oracle's datacenter buildout debt, and Nvidia engaging in deal talks worth three-quarters of a trillion dollars, including guaranteeing OpenAI's debt. Additionally, a Chinese startup released a new competitive AI model, raising questions about the necessity of massive GPU access and data center building. Despite optimism from some investors, the AI market faces potential correction and overbuilding.