Has AI Model Market Begun to Stabilize Around a "Rule of Three" Shape?

At least on mobile devices, ChatGPT remains the share leader, followed by Gemini and then Claude, say analysts at Apptopia. Probably the biggest change is how much share Meta AI has gained, as it now is the fourth-largest model, in terms of mobile device use. 

source: Apptopia 

And it appears the model market is approaching a zero-sum game, where one model’s gains must come from another’s loss, as global usage rates continue to slow. That means a model mostly grows by taking share from another provider. 

And while the market is not yet “stable,” opportunities to dramatically reshape market structure are dwindling. The rule of three appears to be shaping up. 

Capital-intensive industries tend to reach a stable share pattern led by three firms. Just as significantly, the market leader will tend to have twice the share of provider two, which in turn tends to have twice the share of provider three.    

And since share tends to correlate with profit margin, it really matters whether a firm is first or second in a market. Often, the market leader has four times the market share of provider number three. 

By now, if a model is not in the top three, it is very unlikely to break into the top ranks, history might suggest. 

source: Apptopia

Balancing Human Values and AI When Concentrated Market Leadership Will Happen

In principle, it is hard to disagree with Pope Leo XIV, who argues in Magnifica Humanitas that humans values and artificial intelligence must be balanced.

Some critics will complain about AI ownership concentration and outsized market power, to protect human values.

But markets generally develop with a few leaders, whether we like it or not. 

So we still are left with the thorny task of figuring out how to do all that balancing.

Consider similar concerns about the internet. In the late 1980s and early 1990s, many academics, researchers, and early users believed the internet should remain a non-commercial, collaborative environment.

After all, the early internet was a subsidized, academic network encouraging sharing and open exchange. 

The early internet (ARPANET, NSFNET, and connected university networks) was funded almost entirely by governments and research institutions.

This culture produced enduring norms:

• open protocols

• open publication

• free exchange of software

• collaborative development.

• 
  

The turning point came after restrictions on commercial traffic over the NSFNET were lifted in the early 1990s:

• private Internet Service Providers appeared

• domain registration expanded

• browsers made the web accessible

• online retail became feasible

• venture capital entered the industry.

So some worried about:

• commercialization overwhelming academic culture

• advertising degrading user experience

• unequal access

• concentration of economic power.

Concerns about concentration of power will resemble earlier concerns about the internet. 

But the emphasis on “free” might not happen. 

State-of-the-art AI models have substantial ongoing inference costs, so the marginal cost of serving each additional user is not close to zero. And near-zero marginal costs were the enabler for “free” internet services and apps. 

As a result, "everything should be free" is less economically sustainable for AI than it was for web content. 

On the other hand, concerns about concentration of power have already emerged. But it’s a balance. Without the prospect of profit, much less capital would have flowed into software and internet infrastructure, economists will argue.

And it is likely the rule of three will emerge in various segments of the overall AI market, as is true for capital-intensive markets. 

 

source: Mercatus

The rule of three is the idea that in many competitive industries, market structure tends to settle into a small number of dominant firms because scale, fixed costs, and network effects push markets toward concentration rather than endless fragmentation. 

That often leads to a winner takes all market structure.  

In AI, that logic can show up at multiple layers: a few chipmakers can dominate hardware, a few foundation-model providers can dominate models, a few cloud/enterprise ecosystems can dominate platforms, and a few application software vendors can dominate key use cases:

• Hardware. AI chips and the infrastructure around them are capital-intensive, with high fixed costs and strong scale advantages, which makes concentration likely.

• Models. Frontier model development also has steep training costs, data advantages, and distribution effects, so a small set of model leaders can emerge even if many models exist in the long tail.

• Platforms. Cloud and AI distribution layers can become winner-take-most because users gravitate to ecosystems with the best tooling, trust, integrations, and developer gravity.

• Software. Application layers are often more fragmented than infrastructure, but in categories with strong workflow lock-in or standards, the same top-three pattern can appear.

 

Not all Industries feature the rule of three pattern. That can occur when:

•  they have low fixed costs

• weak scale economies

• highly local demand

• strong differentiation. 

Examples include many local services, artisanal goods, custom professional services, and some labor-intensive niches where geography and relationships matter more than national scale. Sectors with rapid product churn and low switching costs can also resist stable three-firm dominance because new entrants can displace incumbents quickly.

It’s hard to see how the various parts of the AI market can avoid developing along a rule of three pattern. 

And that means some critics will be severely disappointed. 

AI Ecosystem "Rule of Three" Coming?

The eventual market structure for the artificial intelligence value chain is a reasonable question, as it was for the internet value chain before it and for virtually every value chain, ever. 

The core question for at least a few possible market leaders: should you own the entire value chain (vertical integration) or dominate a single layer exceptionally well (horizontal specialization)? 

And even for possible market leaders, the idea of becoming “a platform” necessarily entails a horizontal dominance. 

For most firms with less scale, the answer is almost always some form of horizontal specialization. 

Vertical dominance almost always appeals early on, though, as much of the stack does not yet exist, and must be created. 

Early in any market's development, firms face high uncertainty, fragmented or nonexistent supply chains, undefined standards, and limited infrastructure. 

The "value stack" (the full chain of activities from raw inputs to end-customer delivery, including supporting services like logistics, financing, or after-sales) is incomplete or unreliable.

The internet era began with vertically integrated ambitions that mostly failed. Later, many firms prospered by operating “asset-light,” owning as little of the full stack as possible.

The internet's structural lesson might be summarized as “ infrastructure commoditizes and value migrates up the stack.”

The winners were companies that owned the layer closest to the user:

• Google (search/intent)

• Facebook (social graph)

• Salesforce (CRM workflow)

• Microsoft (Office + enterprise identity)

• Amazon (fulfillment + Prime).

Vertical integration seems to appeal most early in value chain development. 

The PC and semiconductor markets were once vertically integrated. 

But, eventually, the supply chains became more horizontal:

• Microsoft for operating systems

• Intel for processors

• Nvidia supplying graphics chips

• Several companies manufacturing hard drives.

The single clearest exception to the "horizontal wins" rule was Apple, which maintained radical vertical integration (silicon → OS → apps → retail).

Given the “early” status of AI, you might guess that vertical approaches are favored by would-be future leaders of the market. 

Very-high infrastructure costs (GPUs, memory, data centers, energy sources) mean that infrastructure costs scale faster than revenues unless you own the stack.

This creates a structural pressure toward vertical integration, largely because high infrastructure costs and scarcity now rearranges infra value. At least for the moment, what cloud was to software as a service, AI infrastructure is forAI and AI agents.

What remains undetermined are the long-term relationships within the value chain. How important will infra remain, and how much differentiation can it provide? How important will vertical integration remain?

Much depends on how today’s bottlenecks are resolved. 

For full-stack integrators (Google, Microsoft, Amazon, OpenAI), bottlenecks in compute, distribution, and enterprise relationships suggest at least significant vertical integration advantages

Long term, “a few” ecosystem winners with significant vertical integration are likely to emerge, with partners occupying key horizontal functions. Applications will likely remain an area where the most specialists will emerge, as has been the case for the internet value chain. 

Internet Layer	Internet Winner	Why They Won	AI Layer	Current Leader(s)	Survivability
Physical infrastructure	Telecom / cable cos (AT&T, Comcast)	Owned the last mile; regulatory moats	GPU compute & data centers	NVIDIA, CoreWeave	Medium — commoditization risk as custom ASICs proliferate; CUDA moat is real but contested
Backbone / routing	Level 3, Cogent (commodity over time)	Traffic volume; peering scale	Cloud hyperscalers (compute fabric)	AWS, Google Cloud, Azure	High for top 3; structural oligopoly with massive switching costs
Horizontal platform / OS	Microsoft Windows, then Android/iOS	Developer lock-in; ecosystem flywheel	Foundation model + API platform	OpenAI, Anthropic, Google DeepMind	Medium-high — differentiation real today, commoditization pressure building
CDN / performance layer	Akamai, then Cloudflare	Edge distribution; hard-to-replicate infra footprint	Inference optimization / edge AI	Cloudflare Workers AI, Groq	High for winners; latency & cost matter enormously at inference scale
Search / intent layer	Google	Owned the demand aggregation point; data flywheel	AI assistant / agent interface	ChatGPT, Perplexity, Google Gemini	Very high — whoever owns the default query interface owns the toll road
Vertical SaaS	Salesforce, Workday, Veeva	Deep workflow + data lock-in in specific domains	Vertical AI (legal, medical, finance)	Harvey (legal), Tempus (oncology), Palantir (gov/defense)	Very high — proprietary domain data + workflow integration = durable moat
Developer tooling / middleware	Twilio, Stripe, Segment	Abstracted complexity; usage-based pricing	AI orchestration & dev tools	LangChain, Weights & Biases, Hugging Face	Medium — commoditization risk as hyperscalers bundle equivalents
Content / media	Netflix, Spotify	Owned the user relationship + proprietary content	AI-native consumer apps	Midjourney, ElevenLabs, Runway	Medium — switching costs low, but brand + proprietary training data matter
E-commerce / marketplace	Amazon, Shopify	Demand aggregation + fulfillment infrastructure	Agentic commerce / AI procurement	Amazon Alexa+, emerging agent platforms	Unknown — biggest open question; whoever controls the purchasing agent controls commerce
"Picks and shovels" enabling layer	Cisco (networking gear), VMware (virtualization)	Sold to all combatants; infrastructure-agnostic	Memory, packaging, power	SK Hynix (HBM), TSMC (fabrication), Eaton (power)	Very high — scarce physical inputs with no software substitute

The internet produced one dominant full-stack integrator per consumer surface (Apple in mobile, Google in search/Android, Amazon in commerce/cloud) and many durable horizontal specialists at layers with genuine switching costs.

AI is likely to produce a similar structure, The full-stack integrators with both infrastructure and consumer/enterprise distribution (Google, Microsoft, Amazon) are best positioned for a role Apple almost uniquely pioneered.

Market dynamics tend to create  a "Rule of Three" (or Rule of Three and Four) structure in mature, stable, competitive markets. 

Bruce Henderson of BCG hypothesized in 1976 that a stable competitive market never has more than three significant (generalist) competitors, with the largest having no more than four times the market share of the smallest, often stabilizing around a 4:2:1 ratio (40-50 percent for the leader : 20-25 percent for number two and 10-12 percent for number three). 

That seems reflected in the internet’s “winner takes most” structure. 

Jagdish Sheth and others validated this across hundreds of industries: three full-line generalists dominate 70-90 percent of the market (by share or profit), while the rest consists of niche specialists (product, geographic, or segment-focused) that thrive on margins rather than volume.

Internet Layer	Internet Winner	Why They Won	AI Layer	Current Leader(s)	Survivability
Physical infrastructure	Telecom / cable cos (AT&T, Comcast)	Owned the last mile; regulatory moats	GPU compute & data centers	NVIDIA, CoreWeave	Medium — commoditization risk as custom ASICs proliferate; CUDA moat is real but contested
Backbone / routing	Level 3, Cogent (commodity over time)	Traffic volume; peering scale	Cloud hyperscalers (compute fabric)	AWS, Google Cloud, Azure	High for top 3; structural oligopoly with massive switching costs
Horizontal platform / OS	Microsoft Windows, then Android/iOS	Developer lock-in; ecosystem flywheel	Foundation model + API platform	OpenAI, Anthropic, Google DeepMind	Medium-high — differentiation real today, commoditization pressure building
CDN / performance layer	Akamai, then Cloudflare	Edge distribution; hard-to-replicate infra footprint	Inference optimization / edge AI	Cloudflare Workers AI, Groq	High for winners; latency & cost matter enormously at inference scale
Search / intent layer	Google	Owned the demand aggregation point; data flywheel	AI assistant / agent interface	ChatGPT, Perplexity, Google Gemini	Very high — whoever owns the default query interface owns the toll road
Vertical SaaS	Salesforce, Workday, Veeva	Deep workflow + data lock-in in specific domains	Vertical AI (legal, medical, finance)	Harvey (legal), Tempus (oncology), Palantir (gov/defense)	Very high — proprietary domain data + workflow integration = durable moat
Developer tooling / middleware	Twilio, Stripe, Segment	Abstracted complexity; usage-based pricing	AI orchestration & dev tools	LangChain, Weights & Biases, Hugging Face	Medium — commoditization risk as hyperscalers bundle equivalents
Content / media	Netflix, Spotify	Owned the user relationship + proprietary content	AI-native consumer apps	Midjourney, ElevenLabs, Runway	Medium — switching costs low, but brand + proprietary training data matter
E-commerce / marketplace	Amazon, Shopify	Demand aggregation + fulfillment infrastructure	Agentic commerce / AI procurement	Amazon Alexa+, emerging agent platforms	Unknown — biggest open question; whoever controls the purchasing agent controls commerce
"Picks and shovels" enabling layer	Cisco (networking gear), VMware (virtualization)	Sold to all combatants; infrastructure-agnostic	Memory, packaging, power	SK Hynix (HBM), TSMC (fabrication), Eaton (power)	Very high — scarce physical inputs with no software substitute

Vertical integration is probably going to work for a few big firms. Most long-term providers in the AI ecosystem will be specialists, though. Most markets ultimately develop that way.