Bubble or Boom?

By early 2026 the AI numbers carried a strange tension: roughly $660–690 billion in capital expenditure chasing about $51 billion in direct revenue, a 10-to-1 gap. An MIT study found that 95 percent of more than 300 enterprise generative-AI initiatives showed no measurable profit impact. Bears reach for the dot-com peak; bulls point to cloud computing circa 2011. Both camps have a case, and that is exactly what makes this one hard.

For founders and operators, this sharpens the most uncomfortable decision in any boom: how to act when the technology is obviously real but the financing around it may not be. It forces you to separate genuine customer demand from budget-cycle spending and fear of missing out, and to think clearly about who is holding which assets when the math finally gets tested.