In July 2025, Yale financial expert Pascual Restrepo published a paper with an intriguing facility: “We Will not Be Missed: Work and Growth in the Age of AGI.” Utilizing economic modeling, Restrepo explores what happens when expert system becomes capable of executing any kind of human task. His verdict overthrows conventional thinking: the economic situation can experience explosive growth while human employees end up being economically unnecessary.
This isn’t a prediction about the future. It’s a theoretical structure for recognizing possibilities that might soon change from scholastic conjecture to useful policy questions.
The Great Economic Split: Bottlenecks vs. Accessories
Restrepo’s model fixate a vital difference. In any type of economic climate, some jobs are traffic jams that constrain general development: energy production, scientific research study, logistics, production, nationwide safety. Other tasks are accessories that add value yet do not constrain financial development: arts, friendliness, therapy, also academic study like economics itself.
Right here’s where synthetic basic knowledge modifications everything. When AGI can convert raw computing power right into effective job, it will naturally focus on traffic jam jobs due to the fact that automating these supplies the most significant economic returns. As computing power increases greatly, AGI systems will manage all the job that in fact drives financial development.
The result is counterintuitive: the economic climate expands at unmatched prices while human payment to that growth approaches zero Human beings might continue working, but just on accessory tasks that don’t drive aggregate economic result, also if they continue to be deeply purposeful to human experience. We come to be financial guests on a machine-driven growth engine.
Explosive Growth, No Human Being Payment
Under Restrepo’s model, economic result becomes figured out nearly completely by available computer power rather than human initiative. If AGI systems handle traffic jam tasks like power production and clinical research, and if computing power proceeds expanding swiftly, financial growth might increase far beyond historic rates.
But this development takes place without humans. Employees’ wages merge toward the price of replicating their deal with makers. Because AGI deals with all financially critical tasks, human labor contributes absolutely nothing important to the production procedure. We might stay used, yet just since utilizing pricey computing sources to change us in accessory functions sets you back greater than paying our earnings.
The mathematical implications are raw: labor’s share of complete economic outcome techniques zero, even as the economic climate itself expands massively. All income streams to those who have computing infrastructure Human workers maintain their jobs yet contribute what Restrepo calls a “vanishingly small” amount to real financial production.
Academic Reception and Real-World Intricacy
Because its presentation at the National Bureau of Economic Research study’s Business economics of Transformative AI conference, the paper has actually created thoughtful conversation within business economics and AI plan circles. MIT’s Neil Thompson praised the modeling technique while noting problems like increasing computing costs that can alter the characteristics.
The version makes several simplifying assumptions. It assumes unlimited computing development, neglects power constraints, and avoids political reactions to such significant economic changes Most importantly, it does not represent just how societies may rearrange when standard economic partnerships collapse.
Possibly much more basically, the model thinks that computational knowledge can perfectly translate into real world capabilities. While AGI could succeed at scientific study, logistics optimization, and other cognitive traffic jams, Moravec’s Paradox recommends that tasks requiring human-like mastery and real-world manipulation continue to be even more tough for equipments Structure and running manufacturing facilities, facilities, and supply chains involves countless physical tasks that might resist automation long after electronic knowledge goes beyond human cognition. This space in between computational power and robot ability can maintain human economic significance in physical duties, also as AGI manages cognitive traffic jams.
These restrictions matter because the version acts as a believing device instead of a crystal ball Its value lies in making clear possibilities and their effects, not in predicting exact end results.
The Paradox of Plentiful Irrelevance
Restrepo’s evaluation reveals an extensive paradox. As AGI automates traffic jam tasks, it can generate extraordinary prosperity while rendering human economic payment outdated. This isn’t traditional technological joblessness where workers require retraining for new duties. Rather, there may be no economically essential duties left for humans to fill up.
The ramifications prolong far past economics. If human work becomes purely accessory while equipments drive all economically essential production, cultures need completely brand-new structures for dispersing the resulting wide range. The paper briefly goes over opportunities like global rewards from compute possession, yet these continue to be theoretical sketches.
A lot more essentially, such a world challenges standard presumptions concerning human purpose and social company. What takes place to definition, identity, and social framework when efficient work no more defines economic connections?
That Possesses the Growth Engine?
If Restrepo’s framework confirms relevant, the critical concern becomes that manages computing facilities. When makers deal with all traffic jam jobs and drive explosive economic growth, ownership of those machines establishes whether prosperity gets shared or concentrated.
The risks of this question are currently playing out in genuine time. OpenAI’s $ 40 billion funding round in March 2025 valued the firm at $ 300 billion, while Anthropic elevated $ 13 billion in September 2025 at a $ 183 billion valuation Meanwhile, Nvidia’s market cap has actually risen past $ 4 3 trillion as need for AI chips explodes. These aren’t just tech investments, they’re bets on owning the facilities that might determine financial winners and losers for generations.
Consider the parallel to oil in the 20 th century Nations that regulated energy sources shaped worldwide economics and geopolitics. Calculate facilities can play a similar duty in the 21 st century. The difference is that oil down payments were geographically distributed by accident of geology. Computer power is being focused by design, with major AI labs controlled by a handful of firms and backed by sovereign wide range funds.
The window for forming these possession frameworks is narrowing as AI capabilities breakthrough. We’re not just watching the next tech boom, we are observing the prospective consolidation of economic power on a scale that can make Criterion Oil look quaint. The choices being made in boardrooms and policy circles today might figure out whether the age of AGI produces shared prosperity or a new form of electronic aristocracy.
Referrals
Acemoglu, Daron, and David Autor. “Abilities, Jobs and Technologies: Implications for Work and Profits.” Handbook of Labor Economics 4 (2011: 1043– 1171
Moravec, Hans. Mind Kid: The Future of Robotic and Human Knowledge Harvard College Press, 1988
Restrepo, Pascual. “We Won’t Be Missed Out On: Job and Growth in the Period of AGI.” In The Economics of Transformative AI University of Chicago Press, 2025 http://www.nber.org/chapters/c 15315
“OpenAI Shuts $ 40 Billion Financing at $ 300 Billion Evaluation.” TechCrunch , March 2025
“Anthropic Increases $ 13 Billion at $ 183 Billion Assessment.” CNBC , September 2025
“Nvidia Market Capitalization Gets To $ 4 3 Trillion.” CompaniesMarketCap , September 2025