Service-Level Disagreement
The article argues that most VC discussion of AI-native services collapses two different businesses into one: Services-as-a-Wedge and Services-as-Delivery-Engine. That mistake distorts expectations around margins, defensibility, TAM, and venture fit.
The Two Models
The article’s main contribution is to separate two superficially similar AI services strategies that have very different downstream economics.
Acknowledge the appeal of outcome-selling
AI services can unlock larger budgets because buyers pay for completed work, not just software access.
Separate wedge from delivery-engine models
The article insists these are different strategies with different prospects for GTM, hiring, margins, and defensibility.
Run the three-question diagnostic
Founders should diagnose vendor position, liability ownership, and how a 10x better model affects competitive dynamics.
Pressure-test margins and TAM
The article closes by challenging assumptions that all labor budgets are realistically capturable or venture-worthy.
Wedge Versus Delivery Engine
The distinction matters because the source of advantage is different in each model, and so is the likely long-run margin profile.
Services-as-a-Wedge
The company enters via outsourced intelligence work, then races to automate the service away before competitors gain the same capabilities.
Services-as-Delivery-Engine
The company keeps human service in the loop because coordination, compliance, and outcome ownership remain essential.
Liability ownership
The article treats this as the clearest dividing line: if you truly own the outcome, your moat can look more like insurance than pure workflow software.
Why The Hype Can Mislead
The article’s skepticism is not about whether AI services can work. It is about how often their economics are being overstated.
Automation trap
Margin expansion may never arrive if humans remain expensive and stubbornly difficult to automate away.
Dispatcher trap
Competitors and incumbents can access the same frontier models, eroding any temporary edge in intelligence-heavy services.
Mirage PMF
Strong growth and retention can look like product-market fit even when true automation and long-run unit economics remain weak.
Venture fundability constraint
Later-stage capital expects software-like growth and software-like margins, which many service-heavy businesses struggle to deliver simultaneously.
Market Examples In The Article
The piece grounds the framework in example verticals to show why some markets fit wedges while others fit delivery engines.
Insurance brokerage
Presented as a wedge category: outsourced, fragmented, and focused on intelligence-heavy comparison and form-completion work.
Transactional legal work
Another wedge example where standardized outputs and existing outsourcing make vendor substitution easier than organizational retooling.
Fund administration
Presented as a delivery-engine market because audits, books, reporting, and system-of-record responsibilities keep human ownership in the loop.
FAQ From The Knowledge Graph
The graph includes linked Question and Answer nodes for the article’s central distinctions and investment critique.