NBER Study Finds "No Productivity Impact" from AI So Far (And Nobody Should be Surprised)

Maybe we should not be surprised that studies of AI productivity often show few results so far. A recent study published by the National Bureau of Economic Research, for example, found:

• around 70 percent of firms actively use AI

• More than 66 percent of top executives regularly use AI, their average use is only 1.5 hours a week, with one quarter reporting no AI use

• firms report little impact of AI over the last three years, with over 80 percent of firms reporting no impact on either employment or productivity. 

source: NBER 

None of that should come as a surprise. Sure, AI adoption is widespread among survey respondents. across the four countries (U.S.; U.K.; Germany; Australia) studied:

• 69 percent of businesses currently use some AI technology

• “text generation” is a top use case

• “visual content creation” follows

• “data processing” is a top-three use case as well.

source: NBER 

But none of those use cases can easily be tied to bottom-line quantitative results very easily, if at all. They should be time savers, but faster text or image creation or some data manipulation, at modest usage rates, in the context of existing business processes, are probably reasonably described as relatively trivial contributors to measurable productivity. 

Also, there is no reason to expect the J curve of technology adoption will fail to be seen here. 

source 

Amara's Law suggests we will overestimate the immediate impact of artificial intelligence but also underestimate the long-term impact. 

Economic historians such as Erik Brynjolfsson and Paul David have documented that transformative, general-purpose technologies tend to follow the J-curve pattern. 

Initial deployment generates negative or flat productivity returns relative to investment, often for a surprisingly long time. 

David's famous 1990 paper on the "dynamo paradox" showed that electrification of US industry began in earnest in the 1880s but didn't produce measurable aggregate productivity gains until the 1920s.

The reasons are structural: firms must reorganize workflows, retrain workers, build complementary infrastructure, and abandon legacy processes before the technology's benefits materialize. 

The productivity gains, when they finally arrive, are real and large, but they accrue after enormous sunk costs and a long gestation period.

And that is going to be a problem for financial analysts and observers who demand an immediate boost in observable firm earnings or revenue, as well as the firms deploying AI that will strive to demonstrate the benefit.

The Great Reversal

Many Christians now are in the liturgical season of Lent, (Feb 18, 2026) to just before Easter, a period for repentance and spiritual renewal. 

Consider, in that regard, the Beatitudes and the Litany of Humility, which both dramatize the great reversal of values called for, where what humans normally prize (status, power, recognition, self‑assertion) is inverted and replaced by poverty of spirit, meekness, and self‑forgetful love. 

source: Laura James 

"The Sermon On The Mount" (Mt 5:3-12) begins with "The Beatitudes," an astounding and jarring reversal of human values, where those who are “blessed” (happy!!) are those one would never assert should be happy: 

• Poor in spirit (humble)

• Those who mourn (their spiritual poverty)

• The meek (gentle out of awareness of their spiritual poverty)

• Who hunger and thirst for righteousness (fervently desire to rectify their spiritual poverty)

• The merciful (forgiving and compassionate)

• The pure in heart (no selfishness)

• The peacemakers (not the "absence of conflict” but the absolute fulfillment of goodness)

• Persecuted for righteousness (misunderstood by the powerful).

Rafael Cardinal Merry del Val’s Litany of Humility illustrates interior movements that illustrate the implications of the Beatitudes. 

source: Friarmusings 

O Jesus, meek and humble of heart, Make my heart like yours.

From self-will, deliver me, O Lord. 

From the desire of being esteemed, deliver me, O Lord.

From the desire of being loved, deliver me, O Lord. 

From the desire of being extolled, deliver me, O Lord. 

From the desire of being honoured, deliver me, O Lord. 

From the desire of being praised, deliver me, O Lord. 

From the desire of being preferred to others, deliver me, O Lord. 

From the desire of being consulted, deliver me, O Lord.

From the desire to be understood, deliver me, O Lord. 

From the desire to be visited, deliver me, O Lord. 

From the fear of being humiliated, deliver me, O Lord. 

From the fear of being despised, deliver me, O Lord. 

From the fear of suffering rebukes, deliver me, O Lord. 

From the fear of being calumniated, deliver me, O Lord. 

From the fear of being forgotten, deliver me, O Lord. 

From the fear of being ridiculed, deliver me, O Lord. 

From the fear of being suspected, deliver me, O Lord. 

From the fear of being wronged, deliver me, O Lord. 

From the fear of being abandoned, deliver me, O Lord. 

From the fear of being refused, deliver me, O Lord 

source: WJLA

 

That others may be loved more than I,  Lord, grant me the grace to desire it.

that others may be esteemed more than I, Lord, grant me the grace to desire it. 

That, in the opinion of the world, others may increase and I may decrease, Lord, grant me the grace to desire it. 

That others may be chosen and I set aside. Lord, grant me the grace to desire it. 

That others may be praised and I go unnoticed, Lord, grant me the grace to desire it. 

That others may be preferred to me in everything, Lord, grant me the grace to desire it. 

That others may become holier than I, provided that I may become as holy as I should, Lord, grant me the grace to desire it. 

At being unknown and poor, Lord, I want to rejoice. 

At being deprived of the natural perfections of body and mind, Lord, I want to rejoice.

When people do not think of me, Lord, I want to rejoice.

When they assign to me the meanest tasks, Lord, I want to rejoice

When they do not even deign to make use of me, Lord, I want to rejoice.

When they never ask my opinion, Lord, I want to rejoice. 

When they leave me at the lowest place, Lord, I want to rejoice. 

When they never compliment me, Lord, I want to rejoice.

When they blame me in season and out of season, Lord, I want to rejoice.

This is an astounding and almost unthinkable “ask,” yet it aligns with the “great reversal” of the Beatitudes, and illustrates the attitudes we might strive for. 

Compared to our typical human values, upside down humbleness and humility in every way.

Value Shifted as Media Morphed from Storytelling to Traffic Monetization

As someone who worked for 40 years in ad-supported media and experienced the harsh realities of the internet, the realities of today’s business are brutal. And that was true before generative artificial intelligence, which is accelerating the underlying economic trends. 

In a nutshell, here’s the business model problem: Many media businesses are no longer primarily “storytelling organizations” but traffic monetization systems. As writers we act as storytellers. But whether that is harnessed financially can often be unclear or unworkable. 

Since advertising “cost per thousand” impressions have collapsed over the last few decades, so have media entity revenues, bringing huge cost pressures to the forefront. 

Platforms such asGoogle, Meta and X have captured distribution and pricing power, driving many former independent or smaller entities out of the market. 

So revenue per article must be less than the cost of human labor to produce that content. So automation becomes a survival move. 

Machines can often produce content abundance, especially when the content has high structure. Humans ideally produce scarcity value when lots of insight, interpretation or “meaning” are required. But generative AI is making inroads there, as well. 

The economics of content production therefore favor using machines to produce mass content at scale, when possible, while humans have the edge only where scarce, specialized or trust-critical content is involved. 

Basically, it is all about marginal cost and associated revenue upside. Investigative reporting might, in some cases, have very-high revenue potential, though it is rare. Original analyses have high production cost, and might have moderate revenue lift. 

Breaking news, data-driven news or automated summaries invariably have low revenue upside. So the choices are fairly simple: automate what does not produce reasonable amounts of revenue, reserving human roles for the more-complicated content that will be a relatively-small part of total content production. 

Content Type	Marginal Cost	Revenue Potential	Economic Role
Investigative reporting	Very high	High but rare	Brand anchor
Original analysis	High	Medium	Subscriber retention
Breaking news rewrite	Medium	Low	Traffic defense
Data-driven updates	Near zero	Low	Volume filler
Automated summaries	~zero	Very low	Search capture

Most user-generated content business models follow a somewhat-similar, but “flipped” pattern. UGC platforms follow the same economic logic (abundance versus scarcity, automation versus humans), but the roles of humans, machines, and “content” are different. 

Abundant, low-value content is automated, while scarce, high-value content is human-produced. 

But users supply the labor for abundance, instead of journalists. The platforms automate selection, amplification, and monetization, so scarcity comes from attention, status, and trust. 

That flows from the differences in media models. Media companies pay to produce content and then monetize the content. 

UGC platforms get their content for free and monetize behavior around that content. In other words, the real product is engagement, not content, per se. So curation is more important than content supply, which is, for all intents and purposes, unlimited. Attention is the source of scarcity. 

Generative AI accelerates the content-creation processes, which becomes even more abundant and lower cost. But the algorithms still curate. So UGC platforms will optimize for watch time; shares; comments or return visits. 

So algorithms will favor emotionally activating content; identity-affirming content or controversial content. That might be likened to “commodity” news, the sort of stuff that, in a professional media context, is structured enough to be automated. 

Top UGC creators, in terms of revenue potential, often provide insights on what matters, what to ignore or how to think about something. They often focus on meaning, which is the same “scarce human” function in professional media.

Ironically, AI increases competition for attention, but raises the premium on human scarcity. That happened with journalism after the internet; music after streaming or photography after smartphones. 

Abundance also tends to make authentic human insight more valuable, not less, even if it remains rare, as surfaced by the algorithms.

AI Might Not be the Actual Cause of Enterprise Job Cuts

Though it might be quite easy to attribute job cuts to the impact of artificial intelligence, that is likely not the actual cause of job cuts in the corporate overhead area at the moment. Many employers arguably “overhired” in the wake of the Covid pandemic. 

During 2020–2022, U.S. firms in technology experienced simultaneous labor scarcity and demand uncertainty from several sources:

• Demand surged suddenly

• Workers exited the labor force or changed jobs at unprecedented rates

• Hiring pipelines broke.

• Losing employees felt existential.

So companies stopped hiring “to plan” and started hiring to secure supply. In essence, labor became something to “stockpile,” not optimize. That was the root of the overhiring wave. 

The rationale was fairly simple:

• Many workers leaned in the direction of  “lifestyle” over “employment”

• Being short engineers, for example, could cap revenue growth

• Being understaffed could cause outages, customer churn, or lost market share.

• Hiring later might be impossible at any price.

So firms rationally decided “it was safer to carry excess labor” than risk being unable to execute.

By mid-2022 to 2023, though, demand normalized. Growth rates slowed; capital became more expensive; productivity fell sharply and the scarcity of labor ended. 

Phase	Labor Market Condition	Company Behavior	Headcount Decision Logic
Pre-Covid (2019)	Balanced labor market	Hire to forecast	Staff aligned to revenue
Early Covid (2020)	Hiring freezes	Defensive posture	Preserve cash
Post-Covid surge (2021)	Extreme labor scarcity	Aggressive hiring	“Hire anyone we can get”
Peak demand panic (2022)	Workers hard to find	Labor hoarding	Hire ahead of demand
Demand normalization (2023)	Labor supply improves	Hiring freezes	Headcount locked in
Correction phase (2023–2024)	Normal job market	Layoffs & restructuring	Resize to revenue
Post-reset (2024–2025)	Rational labor pricing	Targeted hiring	ROI-based staffing

“AI” might be the stated rationale for the job cuts, but that is likely a smokescreen for a past wild hiring binge. AI might eventually have a labor market impact in its own right, but that is likely not what is driving the current wave of big-company layoffs. 

And that “boom and bust” cycle is not terribly unusual for other  industries beyond computing.  

Era	Scarce Resource	Corporate Behavior	Outcome
1999–2000 dot-com	Software engineers	Hire aggressively	Mass layoffs 2001
2006–2007 housing	Construction labor	Expand payrolls	Collapse 2008
2017–2018 oil boom	Petroleum engineers	Labor hoarding	Bust layoffs
2021–2022 tech	Knowledge workers	Over-hire	2023–24 layoffs