Dan Ives Says Slowing AI CapEx Is More Expensive Than Spending More

Rising hyperscaler CapEx is usually modeled as a margin risk.
Dan Ives sees it differently.
His argument is that companies like Microsoft $MSFT cannot afford to slow AI infrastructure spending because the cost of losing strategic position could be far greater than the near-term hit to margins.
Ives compares the AI race to Las Vegas in 1955: there is only one Strip, and every major player wants to secure a place on it before the market matures.
In that framework, CapEx is not simply spending — it is defense against becoming irrelevant.
Aswath Damodaran takes the opposite view.
He argues that hyperscalers still have not clearly explained what business model their massive AI investments are ultimately supposed to produce, leaving investors to question the return on that capital.
Same CapEx. Two completely different interpretations.
Ives sees it as the price of staying on the AI Strip.
Damodaran sees unclear ROI as a warning sign.
$MSFT sits directly inside that debate.