📌 Keynes vs. AI: Should Robots Pay for the Jobs They Destroy?

🔥 Keynes is DEAD. AI KILLED HIM. 🔥

Fernando C. Gaspar

and

Vitor Lima

Feb 14, 2025

Article voiceover

0:00

-15:21

Governments are throwing billions at the economy. It’s not working. Stimulus is fueling stock buybacks, not jobs.

Wages are flat. AI and platform capitalism have hacked the system.

I used to believe Keynes had all the answers. More government spending → More jobs → More growth.

That was before AI companies started making billions without hiring. The money moves up. It never trickles down.

I needed answers. Should we tax AI? Should we abandon Keynesianism? Is there a way out?

Today’s post breaks it down. Why Keynesianism is failing. Why a robot tax won’t save us. What we should do instead.

⚠️ If your job, your business, or your country depends on economic stability—this matters.

👉 Read now, and tell me: Is Keynesianism DEAD, or can it be saved? 🚀

Refer one friend (one is enough) and get the “Global Risk Dashboard” (usually sold by €19,97, see below), an instrument to help you evaluate and manage your exposure to the risks involved in the main transitions taking place in the world.

Refer a friend

📣 Spread the word!

1. Keynes vs. AI

John Maynard Keynes saved economies from collapse. But could his biggest ideas be useless in the age of AI and Algorithm C1apitalism?

Millions of jobs are vanishing. AI doesn’t just automate factory work—it writes, analyzes, and even thinks. Keynes believed government spending could fix unemployment, but what if the jobs never come back?

Some say we need a radical fix: a tax on AI and automation. Others warn this could kill innovation. Who’s right?

2. Why Keynesian Economics Breaks in the Age of AI

Keynesianism assumes that labor is central to economic growth. AI and platform capitalism break that assumption.

Keynesian economics works when demand creates jobs, which then generate income, which then fuels more demand. But what happens when AI allows companies to grow without hiring?

💡 The problem: Tech giants accumulate wealth without spreading it.
💡 The consequence: Wages stagnate, middle-class jobs vanish, and traditional stimulus spending stops working.
💡 The big question: Should we tax AI-driven firms to redistribute economic gains—or is that a mistake?

3. The Collapse of Keynesian Economics in the AI Era

Keynesianism assumes that when demand rises, businesses hire workers to meet it. Wages go up, people spend more, and the economy grows. But AI and platform capitalism break this cycle—companies can now grow without hiring.

AI-driven firms can scale through automation, not human labor. Keynesian stimulus relies on job creation, but when jobs disappear permanently, government spending loses its power to drive economic growth.

3.1. Keynesian Stimulus No Longer Works as Intended

During COVID-19, governments injected $16 trillion into the economy (IMF, 2021). The result?

✅ Stock markets hit record highs.
❌ Wages stagnated (OECD, 2022).
❌ Tech giants absorbed the gains while cutting jobs.

Amazon’s revenue grew 9% in 2023, yet it cut 27,000 jobs (Amazon, 2023). Keynesian stimulus fuels corporate profits, but not employment.

3.2. Platform Capitalism Concentrates Wealth at the Top

Digital monopolies don’t need labor to scale. In 1950, General Motors employed 600,000 workers. Today, OpenAI is valued at $90 billion with fewer than 1,000 employees (SEC Filings, 2023).

Big Tech reinvests in automation, not jobs:

Apple’s revenue hit $394 billion, yet it employs fewer people than Walmart (Apple, 2023).
AI is replacing jobs across industries—375 million workers could be displaced by 2030 (McKinsey, 2017).

3. The Labor-Income-Consumption Cycle Is Breaking

Keynesianism assumes wages drive demand. AI disrupts this by reducing the need for human labor:

📉 Productivity is rising, but wages are not. AI-driven efficiency benefits corporations, not workers (Acemoglu & Restrepo, 2020).
📉 Job markets are polarizing. Middle-income jobs are vanishing, leaving only low-wage gig work or high-skilled tech jobs (Frey & Osborne, 2017).

The richest 1% now control nearly 50% of global wealth, while wages for the bottom 50% have barely moved (World Inequality Report, 2023).

🚨 Keynesianism, in its traditional form, is becoming obsolete. If AI-driven firms don’t hire, government stimulus won’t create jobs—it will only fuel corporate profits. The next section explores potential solutions: should AI be taxed, or is there another way to fix this?

Key Takeaway: Keynesianism Must Adapt, or It Will Become Obsolete

If AI and platform capitalism continue decoupling economic growth from job creation, Keynesianism, in its traditional form, will fail to maintain economic stability. The labor-income-consumption cycle is breaking down, and without intervention, inequality will deepen, demand will stagnate, and economic crises will become more frequent.

Governments cannot rely on traditional stimulus alone. Without policies that ensure AI-driven wealth reaches workers, Keynesian economics will become an outdated relic of the industrial age. The next section explores potential solutions: Should we tax AI, redistribute tech profits, or rethink economic policy altogether?

4. Will an AI Tax Fix the Problem or Make It Worse?

Taxing AI seems like a straightforward Keynesian solution: if automation replaces workers, companies should compensate society by paying a robot tax. The revenue could fund universal basic income (UBI), retraining programs, or public investment in labor-intensive sectors.

But critics argue that taxing AI could backfire—discouraging innovation, driving businesses offshore, and creating bureaucratic nightmares. So, would an AI tax save jobs, or just slow progress?

4.1. The Case for an AI Tax: A Keynesian Lifeline

The core idea behind an AI tax is redistribution—ensuring that the economic benefits of automation aren’t hoarded by tech giants but are reinvested into society.

📌 Preventing wealth concentration: AI-driven firms generate massive profits without hiring workers. An AI tax would ensure that automation benefits the public, not just shareholders (Piketty, 2014).

📌 Funding workforce transitions: Governments could use AI tax revenue to finance reskilling programs for displaced workers (McKinsey, 2017).

📌 Keeping demand alive: If AI eliminates too many jobs, consumer spending collapses. An AI tax could fund UBI or social safety nets, ensuring people still have money to spend (Van Parijs & Vanderborght, 2017).

🚀 Example: South Korea has already implemented a de facto AI tax by reducing tax incentives for companies that automate jobs, aiming to slow labor displacement (South Korean Ministry of Economy and Finance, 2021).

But AI tax critics argue it’s a dangerous move.

4.2. The Case Against an AI Tax: A Job Killer?

🚨 a. It could slow innovation.
Taxing AI might discourage investment in new technologies, making economies less competitive globally (Brynjolfsson & McAfee, 2014).

🚨 b. It’s impossible to define what an “AI job” is.

Should we tax factory robots but not chatbots?
What about self-checkout machines?
Would every spreadsheet replacing an accountant’s job be taxed?

Defining which AI systems are taxable is a logistical nightmare (OECD, 2023).

🚨 c. It could push businesses offshore.
If some countries impose an AI tax while others don’t, companies might relocate to tax-free AI havens, just as corporations already shift profits to low-tax jurisdictions (Zucman, 2019).

🚨 d. It assumes all automation is bad.
History shows that automation doesn’t always eliminate jobs—it shifts them. The Industrial Revolution destroyed some roles but created others (Mokyr et al., 2015). Should we tax progress?

4.3. A Smarter Alternative? The AI Dividend

Instead of punishing automation, some economists propose sharing AI-generated wealth:

✅ AI Wealth Fund – Governments could tax AI-driven profits, not the technology itself, and redistribute gains to society (Atkinson, 2019).

✅ Universal Data Dividends – If AI profits from public data, citizens should receive a data dividend (Lanier, 2018).

✅ Job-Linked AI Taxes – Tax companies that automate jobs without rehiring, while exempting those that reinvest in human workers (Korinek & Stiglitz, 2017).

Bottom line: A robot tax is a blunt instrument—it could work, but smarter solutions exist. The challenge is ensuring that AI benefits everyone, not just billionaires.

5. A Smarter Keynesianism for the AI Era

AI and platform capitalism have rewritten the economic playbook.

Keynesianism, once the ultimate tool for stabilizing economies, is breaking down as automation severs the link between labor and economic growth.

Traditional stimulus spending no longer creates jobs—it boosts corporate profits without trickling down to workers.

So, what’s the solution? Taxing AI isn’t enough. Keynesian economics must evolve.

Instead of treating automation as a problem to be slowed, policymakers must find ways to redistribute AI-driven wealth while fostering innovation.

5.1. The Keynesian Software Update: What Needs to Change?

🚀 a. Tax Profits, Not AI

Instead of a blanket robot tax, governments should implement progressive corporate taxes on AI-driven profits, ensuring that wealth generated by automation benefits society (Zucman, 2019).
The Norwegian Sovereign Wealth Fund provides a model—taxing natural resource profits and reinvesting in the public good. A similar AI Wealth Fund could redistribute automation-driven gains.

🚀 b. Universal Basic Income (UBI) as an AI Safety Net

If AI eliminates jobs faster than new ones emerge, UBI could prevent mass unemployment from collapsing consumer demand (Van Parijs & Vanderborght, 2017).
A data dividend model could fund UBI—tech companies profit from user data, so why shouldn’t individuals get a share? (Lanier, 2018).

🚀 c. AI-Resilient Job Creation

Governments should incentivize industries that require human labor, such as healthcare, education, and clean energy (McKinsey, 2017).
Public AI training programs could help workers transition into higher-skilled roles, ensuring that automation expands opportunities rather than eliminating them.

🚀 d. Rethinking Economic Metrics

GDP growth is outdated. If automation reduces wages, GDP can rise while living standards fall. Policymakers need new economic indicators that account for wealth distribution and job quality (Stiglitz et al., 2018).

5.2. What This Means for You

📌 If you're a worker: Your job might be at risk—not because you’re replaceable, but because automation is changing the economy faster than policies can keep up. The best hedge? Continuous learning and AI-proof skills.

📌 If you're a business leader: AI-driven growth is inevitable, but ignoring the redistribution problem will lead to economic instability. Companies that proactively reinvest in workers will be ahead of regulatory backlash.

📌 If you're a policymaker: Keynesian stimulus needs an update. Taxing AI alone won’t work. A combination of profit-based taxation, UBI, and human-centered economic policies is the only sustainable path forward.

🚨 Keynesianism isn’t dead—it just needs an upgrade. AI isn’t the enemy, but unless we adapt economic policy, automation will create crises instead of progress.

What’s Next?

Next week, we look into “Milton Friedman in the Age of AI and Platform Capitalism”—how his free-market ideology collides with algorithm-driven monopolies and whether his theories still hold up in a world dominated by AI-driven decision-making.

👉 Don’t miss it

I just explained my opinion on this. But I might be wrong. What's yours?

🚀 Should AI profits be taxed to fund public goods, or would that stifle innovation?
🤖 Is Keynesianism outdated in an economy where companies grow without hiring, or can it still be adapted for the AI era?

Drop your thoughts in the comments! 💬 Let’s debate.