<![CDATA[EconWorks]]>https://blog.econworks.comhttps://substackcdn.com/image/fetch/$s_!MxqO!,w_256,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa81687e3-442e-4e1e-9263-f8593f574b19_1024x1024.pngEconWorkshttps://blog.econworks.comSubstackSun, 21 Jun 2026 05:45:37 GMT<![CDATA[The Subsidy Paradox — Why Cursor Sold to SpaceX]]>https://blog.econworks.com/p/the-subsidy-paradox-why-cursor-sold-de0https://blog.econworks.com/p/the-subsidy-paradox-why-cursor-sold-de0Fri, 19 Jun 2026 11:17:55 GMTWe break down the economic structure of the AI developer market in this episode. We discuss how dependence on upstream foundation models compels downstream innovators to flow both capital and highly refined telemetry back to their vertically integrated competitors.


We objectively evaluate the proposed solution, the acquisition of Cursor by SpaceX/xAI. Is the deal a classic case of vertical foreclosure that will harm competition or a necessary defensive merger that ends a data monopoly? We walk through the specific efficiencies of the deal, including cost reduction, performance optimization, and the removal of the innovation barriers.

Read the full article and graphic analysis:

Explore more visual economics content: https://econworks.com

YouTube: https://www.youtube.com/@EconWorks-d3e

Substack:

EconWorks
Industrial organization insights on antitrust, digital platforms, and competition in ecosystem markets.
]]><![CDATA[Control in AI—Episode 2: Governance and Control in AI]]>https://blog.econworks.com/p/control-in-aiepisode-2-governancehttps://blog.econworks.com/p/control-in-aiepisode-2-governanceThu, 18 Jun 2026 11:16:04 GMTAI Antitrust Series. Episode 2: In our second episode, we deconstruct the AI stack (Model → System → Interface) and show how control at the model layer determines access, capabilities, integration, and ultimately who wins the AI economy. We explore why economies of scale, learning effects, and scope make the model layer the main bottleneck. We compare it to the Apple ecosystem and discuss the two competing visions: AI as open infrastructure vs. AI as a tightly controlled platform.
By the end of this episode, you will understand why governance is the new question at the heart of AI—not just who gets to build the best model but who gets to decide the rules of the whole system. It’s perfect for anyone who wants to understand the deeper economics and power dynamics behind the AI revolution.

Read the full article and graphic analysis:

Explore more visual economics content:

https://econworks.com

YouTube:

https://www.youtube.com/@EconWorks-d3e

Substack:

EconWorks
Industrial organization insights on antitrust, digital platforms, and competition in ecosystem markets.
]]><![CDATA[Competition in AI Compute Is About Speed, Not Control]]>https://blog.econworks.com/p/competition-in-ai-compute-is-abouthttps://blog.econworks.com/p/competition-in-ai-compute-is-aboutThu, 18 Jun 2026 11:15:48 GMT

IN 30 SECONDS

AI compute isn’t constrained by chips. It’s constrained by infrastructure.

• Power, permitting, and construction delays limit how fast capacity can be build

• Markets are effectively full → vacancy near zero and prices rising

• Demand continues to surge → The result: persistent shortage

KEY INSIGHT

Competition is not about control.<br>

It is abou…

Read more

]]><![CDATA[Control in AI — Episode 3: Meta, Manus, and the Question of Control in AI]]>https://blog.econworks.com/p/control-in-ai-episode-3-meta-manushttps://blog.econworks.com/p/control-in-ai-episode-3-meta-manusThu, 18 Jun 2026 11:14:18 GMT

📦 In 30 seconds

The Meta–Manus transaction never fully closed.

But it progressed far enough to begin shaping technical direction.

This creates a new category of transaction:
partially realized deals—arrangements that move beyond theory into early integration, yet stop short of full control.

The same ambiguity appears in disputes involving OpenAI.

The key que…

Read more

]]><![CDATA[Why Cursor Sold to SpaceX: The AI Subsidy Paradox]]>https://blog.econworks.com/p/why-cursor-sold-to-spacex-the-aihttps://blog.econworks.com/p/why-cursor-sold-to-spacex-the-aiWed, 17 Jun 2026 11:16:09 GMTWhat if a standalone software company is structurally required to fund and train its biggest competitors?


In this episode of EconWorks, we talk about the “Subsidy Paradox” behind the SpaceX/xAI acquisition of Cursor. We examine the dual nature of the AI coding market, comparing the traditional digital platform paradigm with the new “inverted data siphon” of the AI stack.


We map capital and data flows between downstream interfaces like Cursor and upstream foundation models like Anthropic and OpenAI and analyze the structural vulnerabilities of independent AI developers. We then undertake an objective assessment of the competitive effects of the SpaceX transaction, balancing the traditional theories of vertical foreclosure against the potential efficiencies of defensive integration.

Read the full article and graphic analysis:

Explore more visual economics content:

https://econworks.com

YouTube:

https://www.youtube.com/@EconWorks-d3e

Substack:

EconWorks
Industrial organization insights on antitrust, digital platforms, and competition in ecosystem markets.
]]><![CDATA[Control in AI—Episode 2: When the Model Is the Market]]>https://blog.econworks.com/p/control-in-aiepisode-2-when-the-modelhttps://blog.econworks.com/p/control-in-aiepisode-2-when-the-modelTue, 16 Jun 2026 11:13:06 GMTIn Episode 2 of the Control in AI Series, we deconstruct the AI stack (Model → System → Interface) and explore how control at the model layer dictates access, capabilities, integration, and ultimately who wins the AI economy. We discuss the economies of scale, learning effects, and scope that make the model layer the main bottleneck. We compare it with Apple’s ecosystem and discuss the two competing visions of AI as open infrastructure versus AI as a tightly controlled platform.
By the end of this episode, you’ll know why governance is the key question in AI today—it’s not about who builds the best model; it’s about who gets to decide the rules of the whole system.

Read the full article and graphic analysis:

Explore more visual economics content:

https://econworks.com

YouTube:

https://www.youtube.com/@EconWorks-d3e

Substack:

EconWorks
Industrial organization insights on antitrust, digital platforms, and competition in ecosystem markets.
]]><![CDATA[Are Platforms Breaking Antitrust?]]>https://blog.econworks.com/p/are-platforms-breaking-antitrusthttps://blog.econworks.com/p/are-platforms-breaking-antitrustFri, 12 Jun 2026 11:25:27 GMTWhat happens when the “market” isn’t a product—but a platform?

In this episode, we unpack why traditional antitrust tools struggle with platforms and introduce a new way to think about them. From credit cards to Google to sports leagues, not all platforms work the same way—and that distinction matters more than you think.

Read the full article and graphic analysis:

Explore more visual economics content:

https://econworks.com

YouTube:

https://www.youtube.com/@EconWorks-d3e

Substack:

EconWorks
Industrial organization insights on antitrust, digital platforms, and competition in ecosystem markets.
]]><![CDATA[The Subsidy Paradox: Why Cursor Sold to SpaceX]]>https://blog.econworks.com/p/the-subsidy-paradox-why-cursor-soldhttps://blog.econworks.com/p/the-subsidy-paradox-why-cursor-soldFri, 12 Jun 2026 11:09:48 GMT

In 30 Seconds

  • The Trap: Cursor, the leading AI-native developer interface, caught itself in a structural “subsidy paradox.” By relying on foundation models from OpenAI and Anthropic, Cursor inadvertently funded and trained its own direct competitors. Every enterprise keystroke acted as an inverted data siphon, helping incumbents build rival coding tools.

  • …

Read more

]]><![CDATA[Control in AI—Episode 2: The Question of Control in AI]]>https://blog.econworks.com/p/nextcontrol-in-aiepisode-2-the-questionhttps://blog.econworks.com/p/nextcontrol-in-aiepisode-2-the-questionFri, 12 Jun 2026 11:07:54 GMT

In 30 Seconds
• The Traditional Paradigm: Prevailing antitrust frameworks assume that control materializes upon the formal execution of an acquisition.
• The AI Reality: In AI markets, de facto control frequently emerges long before legal ownership transfers.
• The Broader Issue: Influence accrues incrementally via technical alignment, compute infrastructu…

Read more

]]><![CDATA[Mattresses, Mergers & Vertical Power: AI Wraps Up Season 1]]>https://blog.econworks.com/p/mattresses-mergers-and-vertical-powerhttps://blog.econworks.com/p/mattresses-mergers-and-vertical-powerThu, 11 Jun 2026 11:15:56 GMTThe season finale tackles the FTC’s challenge to Tempur Sealy’s vertical merger with Mattress Firm. Was the merger a smart efficiency play in retail and manufacturing, or did it risk anticompetitive foreclosure?

In this episode, AI critiques the expert economic analysis, focusing on vertical merger guidelines, concentration risks, and real-world retail dynamics. We compare the AI take to the court’s perspective and close season 1 with big-picture lessons on how antitrust should handle innovation, efficiency, and evidence in fast-changing markets.

Read the full article and the graphic analysis.

Explore more visual economics content:

https://econworks.com

YouTube:

https://www.youtube.com/@EconWorks-d3e

Substack:

EconWorks
Industrial organization insights on antitrust, digital platforms, and competition in ecosystem markets.

]]><![CDATA[When the Platform Is the Market—Rethinking Antitrust]]>https://blog.econworks.com/p/when-the-platform-is-the-marketrethinkinghttps://blog.econworks.com/p/when-the-platform-is-the-marketrethinkingWed, 10 Jun 2026 11:23:13 GMTAntitrust law begins with a straightforward question: what alternatives do consumers actually have?

That question becomes much harder—and sometimes misleading—when applied to platforms.

In this episode, we introduce a new framework for understanding platform market definition. Not all platforms operate the same way, and treating them as if they do leads to flawed analysis and bad outcomes.

We break platforms into three distinct types:

  • Transactional platforms—where both sides meet in a single exchange

  • Non-transactional platforms—where interactions happen over time

  • Regulatory platforms—where the platform controls who gets to compete

Using real-world examples—including sports leagues and the NASL case—we show why existing antitrust tools often fail and what a better approach might look like.

This is the first piece in a series developing a new framework for platform economics and market definition.

Read the full article and the graphic analysis.

Explore more visual economics content:

https://econworks.com

YouTube:

https://www.youtube.com/@EconWorks-d3e

Substack:

EconWorks
Industrial organization insights on antitrust, digital platforms, and competition in ecosystem markets.
]]><![CDATA[Main Title: Mattress Merger Wars: Vertical Efficiencies or Anticompetitive Risk? AI Verdict]]>https://blog.econworks.com/p/main-title-mattress-merger-wars-verticalhttps://blog.econworks.com/p/main-title-mattress-merger-wars-verticalTue, 09 Jun 2026 11:16:23 GMTThe FTC challenged Tempur Sealy’s acquisition on vertical and horizontal grounds. In the season finale, AI breaks down the expert report, focusing on efficiencies, foreclosure risks, and real retail dynamics. We compare the critique to the court’s perspective and wrap up Season 1 with key lessons for future antitrust cases.

Read the full article and graphic analysis:

Explore more visual economics content:

https://econworks.com

YouTube:

https://www.youtube.com/@EconWorks-d3e

Substack:

EconWorks
Industrial organization insights on antitrust, digital platforms, and competition in ecosystem markets.
]]><![CDATA[When the Platform Is the Market]]>https://blog.econworks.com/p/when-the-platform-is-the-markethttps://blog.econworks.com/p/when-the-platform-is-the-marketSat, 06 Jun 2026 20:22:34 GMT

Antitrust market definition sounds technical. It isn’t.

At its core, it asks a simple question: what alternatives do customers actually have?

For ordinary products, the answer is usually straightforward. If the price of one product rises, consumers switch to another. That substitution defines the market.

Platforms complicate this picture. And in some cases…

Read more

]]><![CDATA[AI v. Judge — Episode 5: FTC vs. Tempur Sealy*]]>https://blog.econworks.com/p/ai-v-judge-episode-5-ftc-vs-tempurhttps://blog.econworks.com/p/ai-v-judge-episode-5-ftc-vs-tempurSat, 06 Jun 2026 20:12:51 GMT

Key Takeaways

  • The FTC challenged the vertical merger of Tempur Sealy (one of the world’s largest mattress manufacturers) and Mattress Firm (the nation’s largest bedding retailer), arguing that the merged firm could foreclose rival mattress suppliers through control of retail distribution.

  • The plaintiff’s economic analysis advanced a narrow “premium” mattr…

Read more

]]><![CDATA[The $2.5 Billion Question in Pharmaceutical Antitrust]]>https://blog.econworks.com/p/the-25-billion-question-in-pharmaceuticalhttps://blog.econworks.com/p/the-25-billion-question-in-pharmaceuticalFri, 05 Jun 2026 11:11:24 GMTIn this episode, we examine the economic foundations of reverse-payment antitrust law through the lens of the Takeda verdict.

The Supreme Court’s *Actavis* decision shifted the focus from patent validity to payment size—but left unresolved how courts should measure what counts as “large.”

We explore the circularity embedded in the framework, the role of damages models, and why jury decisions may ultimately substitute for missing economic methods.

Read the full article and the graphic analysis:

Explore more visual economics content:

EconWorks

YouTube

Substack

]]><![CDATA[Google’s Ad Tech Empire Under the Microscope: AI Antitrust Critique]]>https://blog.econworks.com/p/googles-ad-tech-empire-under-thehttps://blog.econworks.com/p/googles-ad-tech-empire-under-theThu, 04 Jun 2026 11:18:52 GMTIn this episode, we talk about the complexities of digital advertising in the US. DOJ v. Google ad tech case. Google’s integrated platform dominates ad servers, exchanges, and networks, and critics have accused it of excluding rivals.

The AI looks at the main expert arguments about auctions, vertical integration, and market power and compares the criticism to the trial evidence. In this episode we unpack important but technical concepts like header bidding and multi-homing and put them in plain English. We also take a new look at one of the biggest ongoing antitrust battles in tech.

Read the full article and the graphic analysis:

Explore more visual economics content:

EconWorks

YouTube

Substack

]]><![CDATA[This Antitrust Rule Doesn’t Work — Here’s Why]]>https://blog.econworks.com/p/this-antitrust-rule-doesnt-work-hereshttps://blog.econworks.com/p/this-antitrust-rule-doesnt-work-heresWed, 03 Jun 2026 11:04:47 GMTIn FTC v. Actavis, the Supreme Court held that large reverse payments could be evidence of anticompetitive conduct. But how do courts determine that a payment is “large”?

The answer involves the probability that a generic would have won the patent case, a number for which there is no generally accepted way of estimation in antitrust litigation.


In this episode, using the recent Takeda case of an $885 million verdict as a case study, it explains:

  • Economic logic of reverse payments

  • Why is the Actavis framework circular?

  • How damages models rely on unspoken assumptions

  • Why juries prefer stories to methods

This is not only a case about pharmaceuticals but also a structural problem in antitrust economics.

Read the full article and the graphic analysis:

Explore more visual economics content:

EconWorks

YouTube

Substack

]]><![CDATA[Google Ad Tech Monopoly? AI Critiques the DOJ’s Expert Arguments]]>https://blog.econworks.com/p/google-ad-tech-monopoly-ai-critiqueshttps://blog.econworks.com/p/google-ad-tech-monopoly-ai-critiquesTue, 02 Jun 2026 11:05:56 GMTExperts debated whether Google’s integrated platform provided efficiency or illegal foreclosure in the high-stakes US DOJ v. Google ad tech case. In Episode 4, the analysis uses AI to critique the key economic claims and compare the evidence and the outcomes of the trial. This is a deep dive into complex digital marketplaces.

Read the full article and the graphic analysis:

Explore more visual economics content:

EconWorks

YouTube

Substack

]]><![CDATA[The $885 Million Question: Actavis and the Circularity Problem]]>https://blog.econworks.com/p/the-885-million-question-actavishttps://blog.econworks.com/p/the-885-million-question-actavisFri, 29 May 2026 13:59:22 GMT

In 30 Seconds

  • The Supreme Court in FTC v. Actavis held that large payments signal weak patents

  • But to determine whether a payment is “large,” courts must already know patent strength

  • That creates a circular liability test with no accepted methodology

  • In Takeda, the jury resolved this gap by selecting a but-for entry date from competing scenarios

  • The result: …

Read more

]]><![CDATA[AI v. Judge — Episode 4: United States v. Google (Ad Tech)]]>https://blog.econworks.com/p/ai-v-judge-episode-4-united-stateshttps://blog.econworks.com/p/ai-v-judge-episode-4-united-statesFri, 29 May 2026 13:55:28 GMT

Key Takeaways

  • The DOJ challenged Google’s conduct in open-web display advertising, arguing that Google used its dominant publisher ad server (used by more than 90% of major publishers) to tie its ad server to its own ad exchange (AdX) and foreclose rival ad tech platforms through self-preferencing rules, data advantages, and feature restrictions.

  • The DOJ’…

Read more

]]>