Google DeepMind published an essay on 29 October 2026, “The Economics of Intelligence,” by Alex Imas, Andrew Koh and Shane Legg, which says “the individual AI agent is the wrong economic object”.
The essay sorts AI progress into four regimes and says that “because wage effects arrive before organizational restructuring, macro statistics of employment rates provide poor evidence that AI is not having an effect on the economy”. It draws on a 108-page paper, “Organizing Intelligence,” released the same day.
Short explainer · DeepMind economics of AGI
A vertical news Short narrated from the headlines and opening lines of this piece and four related pieces.
Business Pill · AI Agents
An evergreen Business Engineer explainer on AI agents. It teaches the general idea only and does not discuss DeepMind’s essay.
The key insight: As we read it, the essay’s test for AI’s economic impact is not whether one model beats one person, but whether an organization of agents beats an organization of people, and it says the first place that shows up is wages.
The essay says economic transformation depends on when two capability thresholds are crossed. Competence, gains in individual skills like math, writing or coding, “will smoothly shift wages, prices and the mix of humans and AI agents in a firm while leaving industrystructure largely intact”.
Coordination is the ability of agents to work together inside a firm. The essay says “Once coordination reaches a critical threshold, it triggers large, abrupt jumps in market concentration, employment, and wages”.
Imas wrote on X that he joined Google DeepMind to develop a research program on the economics of AGI, and that this is the first in a series of projects from the group.
The four capability regimes in Google DeepMind’s essay, laid out on its two dimensions, competence and coordination; the quoted phrases are the essay’s own. Source: Google DeepMind, The Economics of Intelligence, 29 October 2026.
What Was Posted on X
I joined @GoogleDeepMind to develop a research program on the economics of AGI. Today, together with Andrew Koh (@andrewjkoh) and Shane Legg (@ShaneLegg), we are incredibly excited to release The Economics of Intelligence, the first in a series of projects from our group on how AGI will change the economy.
How do capability improvements transform the economy, and when In this project, we define thresholds of economic capabilities, e.g., minimal AGI, based on how they transform the economy. We then track how the economyโmarket structure, employment, and wagesโchanges during the transition across capability regimes.
We have seen exponential increases in what AI agents can do but the broader economy seems almost remarkably unchanged. Part of the reason is that the solo AI agent is the wrong economic object; economic transformation is an equilibrium outcome where organizations compete. Looking at how intelligence functions within organizations, transformation will depend on when two key capability thresholds are crossed: competence (gains on individual skills, e.g., coding, math) and coordination (gains on ability of agents to cooperate and work together).
We show that what economic transformation looks like depends on whether AI becomes supercompetent, supercooperative, or bothโthe regime of artificial superintelligence (ASI). Tracking this transition shows why the real visible transformation may come from improvements in coordinationโa capability most benchmarks donโt yet track. Past a certain threshold, AI capabilities can break the Coasian ceiling on firm size and lead to an economy that looks drastically different than the one we see today.
Two main takeaways: 1. To understand the path of transformation, we should be tracking agentic coordination more closely. 2. The impact of AI will likely show up in wages before market structure or employment.
We are really excited about this paper. Excited to hear peopleโs thoughts.
Alex Imas on X, 29 Oct: he joined Google DeepMind to develop a research program on the economics of AGI, and with Andrew Koh and Shane Legg released The Economics of Intelligence, the first in a series of projects from the group.
New Google DeepMind paper says AI is already changing the economy, just not in the way people look for.
Its changing through leaner startups, flatter org charts and slower hiring for some entry-level jobs, but not through mass job losses.
The paper argues that big economic change from AI depends on agents learning to coordinate, not on each model getting smarter.
AI models already match experts on many solo tasks, yet jobs and productivity numbers have barely moved. Most benchmarks test 1 model alone, but firms compete as teams.
They built an economic model that tracks 2 abilities separately: competence on single tasks and coordination between agents. More competence moves wages gradually, but coordination past a tipping point lets a few large AI firms suddenly take over big, standardized markets.
In a quiz of 100 people, pairs matched to cover each other’s weak spots beat random pairs by 32%. AI copies share the same weak spots, so a model that beats every person alone can still lose to a well-matched human team.
Rohan Paul (@rohanpaul_ai) summarises the Google DeepMind paper on X. He writes that AI is changing the economy through leaner startups, flatter org charts and slower hiring for some entry-level jobs, not through mass job losses, and that in a quiz of 100 people, pairs matched to cover each other’s weak spots beat random pairs by 32%. Those are his summary of the paper’s findings.
The Four Regimes
The essay names the regimes minimal AGI, supercompetent AI, supercooperative AI and artificial superintelligence. Under minimal AGI, where AI matches or exceeds average human capability across tasks, it says “almost nothing structurally changes in the economy” and “All of the changes occur in wages”.
Under supercompetent AI, where AI outperforms every human expert except at coordination, the essay describes “abundance without concentration”. Under supercooperative AI, it says “the economy experiences a sudden phase transition” and returns to scale “flip from being concave to convex”. Under artificial superintelligence, it says smaller niche markets no longer shelter human teams.
Why the Economy Still Looks the Same
The essay says that “headcounts look very similar to a few years ago, and aggregate productivity numbers have barely budged”. Its explanation is that “AI models are getting exponentially better in individual competence while lagging behind in coordination”.
It reports that the authors gave 100 people a quiz across six domains and that “When participants are paired to complement each other to solve a specific task, they outperform random pairs by 32%”. It also quotes OpenAI’s Noam Brown saying he “wouldn’t even attribute 10% of the credit” to multi-agent capabilities over single-agent performance.
Where the Data Sat on 29 October
The same day, the Bureau of Economic Analysis put US third-quarter growth at 3.4%; by our arithmetic from BEA’s contribution table, information processing equipment and software added 1.09 points, about 32% of it, as reported in our GDP piece.
On ServiceNow’s 28 October call, CFO Gina Mastantuono said, according to a transcript published by Platform Aeronaut, that “efficiency gains means that we’re able to not hire as many people”. Microsoft’s CFO Amy Hood said its second-quarter cost guidance “includes an estimate for the recently announced employee voluntary exit offer”.
The essay itself names wages and job roles as the signals to watch: “wages and job rolesโnot headline GDP or industry concentrationโremain the earliest signals of change”.
The Structural Read
As we read the framework, competence and coordination do different jobs: the essay says competence sets the level of output per worker, while coordination decides how big a firm can grow before it becomes too messy to function.
As we read its account of today, the essay’s explanation for flat headcounts is a coordination gap: it quotes OpenAI’s Noam Brown saying swarms of agents struggle to divide labor productively.
As we read its measurement section, the essay asks for benchmarks of organizations of agents and for statistics that track both payroll and compute spending, alongside today’s solo-agent leaderboards.
Google DeepMind, The Economics of Intelligence, 29 Oct 2026
“Separating competence and coordination yields an important insight about economic transformation: competence reshapes the economy gradually, whereas coordination abruptly transforms it.”
Three Implications
FOR EMPLOYERS The essay says roles flip inside lean firms under supercompetent AI: pay for routine specialist tasks gradually declines, while human coordinators who set direction multiply the output of a team of agents.
FOR INVESTORS The essay says that under supercooperative AI, highly synchronized AI organizations in deep, standardized markets can grow large and then suddenly outcompete smaller firms, while niche markets with bounded demand behave differently.
FOR WHAT TO WATCH The essay names wages and job roles, not headline GDP or industry concentration, as the earliest signals, and calls for benchmarks that test how multi-agent systems divide problems and share context.
The framework puts it in two lines: “The first is what your underlying technology can do. The second is what your users can actually do with it.”
As we read it, DeepMind’s essay describes a similar gap at the level of the firm: capability that individual agents have, and coordination that organizations cannot yet use. The framework was published in June 2026 and does not discuss DeepMind’s essay.
The essay is a framework and a model, not a forecast with dates; it does not say when the supercooperative threshold will be crossed or give a measured value for it. It does not discuss ServiceNow, Microsoft or the third-quarter GDP release; placing them beside its argument is this piece’s choice. This piece rests on the DeepMind essay and the paper’s abstract, both dated 29 October 2026, and Imas’s post on X.
This Week In The AI Supercycle
Execution Fault Lines, Concentrated Demand, and Geopolitical Firewalls
This week’s issue: the AI build is not short of demand for its parts. It is short of cheap money, and its end revenue rests on a narrow base of large, often venture-funded buyers.
Google DeepMind’s economists say AI’s first visible effect is on wages, and that larger changes in firm size and employment wait on agents learning to coordinate, which they say most benchmarks do not yet measure.
A note on sourcing. DeepMind’s statements come from its essay “The Economics of Intelligence” and the paper “Organizing Intelligence,” both dated 29 October 2026 and read that day; Alex Imas’s post on X is context. The BEA, ServiceNow and Microsoft passages come from this publication’s earlier pieces, linked in the text.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
