Glonce

Two-thirds of IT leaders report AI results, but few would interrupt the CEO's vacation over them

businessThe Glonce editorsThe Decoder🔥53Glonce rating53Impact45Novelty35Relevance85Surprise40Players35Credibility70
Two-thirds of IT leaders report AI results, but few would interrupt the CEO's vacation over them

The AI bubble debate keeps returning to one question: is AI paying off fast enough to justify the money being poured into data centers and chips? Hard, economy-wide numbers on that return are scarce, so the debate leans heavily on anecdotes. Azeem Azhar, a British tech entrepreneur and founder of the research group Exponential View, offered one such anecdote in a podcast conversation with Nicholas Thompson of The Atlantic. Azhar described speaking to about 160 IT vice presidents in Las Vegas. He first asked who could point to measurable AI results, and asked them to stay standing if so; two-thirds remained on their feet, more than he expected. He then asked a harder question: who had results good enough to interrupt the CEO's summer vacation. Only about eight people stayed standing. Azhar's takeaway is that companies are making progress with AI, but slowly, and whether that pace justifies current investment levels remains an open question. He noted that some executives report their boards growing more ambitious after early wins, and that even in slower markets such as Italy, CEOs told him trust is building and budgets are rising despite missteps. At the same time, he pointed out that many companies are shifting from expensive frontier models toward open-weight alternatives, a shift that could let AI usage keep growing while still leaving the buildout underfunded if not enough revenue flows back to pay for it. Azhar also raised what he called the "bear version of the story": a Boston Consulting Group survey found that about 70 percent of CEOs worldwide say AI success matters for how they are perceived in their roles, which gives them an incentive to describe results more favorably than reality supports. Azhar said he has no simple answer to the AI bubble question himself, calling the situation "finely balanced."

Key facts

  • Azeem Azhar asked about 160 IT vice presidents in Las Vegas if they could point to measurable AI results; two-thirds stayed standing.
  • When asked whose results were good enough to interrupt the CEO's summer vacation, only about eight of the roughly 160 remained standing.
  • Azhar's conclusion: companies are progressing with AI, but slowly, and whether that pace justifies current investment levels is unresolved.
  • A Boston Consulting Group survey found about 70 percent of CEOs worldwide say AI success affects how they are perceived, giving them incentive to overstate results.
  • Azhar noted many companies are shifting from expensive frontier models to open-weight alternatives, which could grow AI usage without generating enough revenue to fund the buildout.

Why it matters

The story is a data point, however anecdotal, in the wider argument over whether AI investment is justified by real returns. The gap Azhar describes, most IT leaders see some measurable result but few see anything that rises to board-level urgency, captures the murkiness at the center of the AI bubble debate: usage and small wins exist, but nothing an economy-wide metric can yet confirm as sufficient to fund the current scale of spending on chips and data centers.

Who it affects

IT vice presidents and other technology executives who report AI progress upward, the CEOs and boards who decide whether to keep raising AI budgets, and anyone trying to gauge whether current AI infrastructure spending is backed by real revenue.

How to use it

There is no product or service here; the value is as a reference point when weighing claims of AI ROI. Readers assessing corporate AI progress reports can note Azhar's distinction between having a measurable result and having one significant enough to escalate to leadership, and the incentive problem the Boston Consulting Group survey points to.

How solid is it

The account comes from Azeem Azhar's own description, given in a podcast conversation with Nicholas Thompson of The Atlantic, of an informal show-of-hands exercise with a self-selected room of roughly 160 IT vice presidents in Las Vegas; no date, location context, or company breakdown beyond that is given. It is anecdotal by nature, not a rigorous survey, though the article also cites a separate, named source, a Boston Consulting Group survey, for the figure on CEOs' incentives to overstate AI success.

Risks and caveats

A show of hands at one conference is not a statistically representative sample, and the 160 attendees' identities, industries, and countries are not specified beyond being met in Las Vegas. Azhar himself frames the overall picture as "finely balanced" and offers no firm answer on whether AI's return on investment justifies its cost; the piece also flags that CEOs have a documented incentive, per the Boston Consulting Group survey, to describe AI results more favorably than warranted.

“the situation is "finely balanced."”

— Azeem Azhar