Don't mistake AI adoption for lasting productivity gains

Loose money plus AI does not automatically equal a durable productivity boom. The article makes that leap: easy liquidity, solid growth, and faster tech adoption should, by logic, lift output per worker. Let's be real: that logic has holes you can't patch with acronyms and optimism.

Start with what the piece gets right. Having loose money, strong growth, and a plausible productivity engine is about as good as macro backdrops get. If you’re an equity investor, that’s the trifecta you dream about: financing is cheap, demand is there, and management teams have a story that sounds like “future margins go up.” On that narrow, market-facing horizon, the Livewire argument holds.

But markets and economies don’t live on the same timeline.

Loose money and strong growth are not independent dials you can keep turned up indefinitely. Central banks don’t sit on their hands if inflation re-ignites or asset prices start to detach from reality. Credit can look easy while imbalances quietly build in housing, private markets, or corporate balance sheets; then it shuts abruptly when sentiment flips. Treating liquidity as a structural tailwind, rather than a cyclical policy stance with political constraints, is less analysis and more wish-casting.

AI is the second leap of faith. The column treats adoption as if it were a clean transmission mechanism: install AI, get productivity. In practice, AI can affect productivity through three broad channels: substituting routine tasks, augmenting complex work, and enabling new products or services. Those play out on different timelines. Augmentation can lift output per worker if people actually change how they work; substitution might just cut headcount without moving the output-per-hour needle; new products can take years before national accounts pick up anything at all. The article waves past those lags as if timing doesn’t matter to the story.

Then there’s the gap between adoption and diffusion. A handful of firms can move fast—especially those already deep in software and cloud. That doesn’t mean the median manufacturer, logistics operator, or professional-services shop gets the same uplift. Historically, that diffusion gap has been measured in years, sometimes decades. The math doesn't lie: early adopters can reap big gains while the rest of the economy looks disappointingly familiar on any productivity chart.

Corporate behavior is the hinge here. Loose money can finance capex, retraining, and systems integration—or it can underwrite buybacks, debt-funded acquisitions, and “AI strategy” slideware that never leaves the boardroom. The original article puts a lot of faith in the capital-allocation channel without asking the uncomfortable question: how many management teams are really prepared to tear up workflows, rewrite incentives, and absorb the short-term disruption that comes with genuine process change?

You don’t boost productivity by stapling a chatbot onto an existing bureaucracy.

History is not exactly on the side of clean, linear tech narratives. Think about the IT boom: trillions spent on hardware and software, stock prices flying—and the famous “productivity paradox” when economists went looking for the payoff. The gains eventually came, but they were uneven and often concentrated in firms that fundamentally reorganized how they operated, not just those that bought the latest tools. AI is likely to rhyme with that pattern more than it resembles a smooth, economy-wide lift.

The distribution question is where the optimism really starts to fray. Productivity gains clustered in a small set of firms or regions can push equity indices higher without moving median wages or broad consumption much. That matters because sustaining strong growth under loose monetary conditions depends on broad demand and investment, not just on a higher multiple for the winners. The article’s “what more do you want?” framing ignores who actually gets the upside and who gets disruption without a clear path to share in the gains.

Politics will not stay offstage. If AI adoption is perceived as amplifying inequality or wiping out specific job categories, policy will react—through regulation, taxation, labor rules, or competition policy. Those responses won’t be neatly “pro-productivity” or “anti-productivity”; they’ll be messy and path-dependent, shaped by headlines rather than white papers. The original piece treats policy as static scenery. It isn’t. It’s a feedback loop.

From my old seat watching traders latch onto narratives, I learned a simple rule: when the story sounds this clean—liquidity, growth, tech, productivity—check what’s missing. What’s missing here is a serious look at sequencing risk. You can have a phase where loose money and AI hype inflate asset prices and encourage investment stories long before the hard operational work shows up in the data. If that window closes—because policy tightens, politics shifts, or investors lose patience—you can be left with a half-finished transformation and a lot of sunk optimism.

So yes, the ingredients the article lists are unusually favorable, and it’s understandable that Livewire Markets leans into that. But without the unglamorous follow-through—management discipline, painful reallocation, and policy that survives the first backlash—this “perfect setup” for productivity more likely turns into another cycle where markets price in the win long before the economy actually earns it.