RIC Framework Navigation
This framework is presented as a core definition with related applications and references.
Framework
Updated 2026
RIC Index
Updated 2026
Updated 2026
Tracker
Coming soon
Reality Infrastructure Companies (RICs)
A framework for understanding where power compounds in the AI era
Purpl uses the term Reality Infrastructure Companies (RICs) to describe a small class of firms that operate at a deeper layer than traditional technology companies.
They are not defined by products, apps, or software categories.
They are defined by the layer of reality they control and by how artificial intelligence has become the control system operating on that layer.
The purpose of this framework is not prediction but clarity.
-
A Reality Infrastructure Company (RIC) owns an irreplaceable, live layer of reality and deploys first-party AI to observe, interpret, and act on that layer in real time.
AI is not the product.
AI is the operator.
-
As AI capabilities have advanced, existing categories have started to break.
“AI companies” describe tools, not power
“Platforms” describe distribution, not control
“Big Tech” describes size, not structure
What actually differentiates the most durable companies today is not model quality, feature velocity, or quarterly execution.
It is structural proximity to reality.
RICs sit where:
signals are generated
decisions are made
actions propagate back into the world
This framework exists to describe that structure clearly.
-
The RIC framework exists to prevent a common analytical error in the AI era:
mistaking tools for control.
It prevents over-indexing on models while under-weighting the layers those models operate on.
It prevents sector-based thinking in a world where power compounds across layers, not industries.
And it prevents confusing rapid innovation with durable advantage.
RICs are not defined by what they build — but by what they operate.
-
A company qualifies as a Reality Infrastructure Company only if it meets all four conditions below.
1. Ownership of a live, non-replicable reality layer
The company controls data that is:
continuously updating
structurally tied to real-world activity
impossible to recreate externally
Examples include search intent, attention, discourse, commerce flows, or physical-world sensing.
2. Embedded distribution
The AI is not an optional feature or standalone tool.
It is embedded in default behavior:
how people search
how they communicate
how they transact
how they move through the world
3. Closed feedback loops
RICs operate closed loops where:
action → data → model → action
This creates compounding control, not incremental optimization.
4. First-party AI sovereignty
The company develops and controls its own core AI systems that operate the primary loop.
RICs do not depend on rented intelligence for the function that defines them.
-
Reality Infrastructure Companies operate a closed loop:
Reality → Data generation → AI control system → Action on reality → (repeat)
The tighter and more continuous this loop, the more durable the control.
-
RICs are not:
AI wrappers
SaaS tools using third-party models
Companies with data but no control loop
Firms where AI is additive rather than operational
Many successful companies use AI.
Very few run reality through it.
-
This term and framework were first articulated by Purpl in 2026 to describe a structural pattern existing categories failed to capture.
Reality Infrastructure Companies (RICs) is not an established industry term.
It was introduced to describe a structural pattern that existing language proved insufficient to explain.
If the term is adopted, it will be because it proves useful.
If it is not, the underlying structure remains unchanged.
Frameworks exist to clarify reality — not to own it.
-
Some companies sit close to RIC qualification without fully meeting the criteria.
Apple controls devices and distribution, but AI is not yet the primary control system operating the core loop.
OpenAI and similar labs control intelligence, but do not own a proprietary reality layer.
Many AI-first startups build powerful tools but operate downstream of infrastructure they do not control.
The distinction is structural, not qualitative.
-
The RIC Index is an application of this framework to publicly investable companies.
It is illustrative, not definitive.
The framework stands independently of any specific index construction.
Within Purpl, the RIC framework is used to:
analyze long-term compounding power
compare companies across traditional sectors
reason about AI beyond models and benchmarks
The framework is designed for multi-decade structural analysis, not short-term forecasting or quarterly performance attribution.
-
Framework introduced: 2026
Revision cadence: As needed, when reality changes
RIC Index
Application of the Reality Infrastructure Companies framework
The RIC Index applies the Reality Infrastructure Companies (RICs) framework to publicly investable companies.
It is intended as an illustration of how the framework operates when applied to real markets, not as a definitive or exhaustive list.
Inclusion criteria
A company is included in the RIC Index only if it:
Meets all four RIC qualification rules
Is publicly listed and investable
Controls a live, non-replicable layer of reality
Operates first-party AI as part of the core control loop
The index is reviewed periodically and updated only when structural conditions change.
Review cadence
Index reviewed periodically
Constituents change only upon clear structural shifts
No short-term optimization or performance targeting
Scope and limitations
The RIC Index is illustrative, not definitive.
It represents one application of the RIC framework under current market conditions.
The framework itself stands independently of any specific index construction.
Status:
RIC Index introduced: 2026
Maintained by: Purpl
Current RIC Index constituents
-
Reality layer: Global knowledge access and search intent
-
Reality layer: Identity, social graph, and attention
-
Reality layer: Commerce, logistics, and economic flow
-
Reality layer: Enterprise workflows and operating environments
-
Reality layer: Physical-world sensing and movement
Structurally eligible but private
-
Reality layer: real-time public discourse and narrative formation
-
Reality layer: global attention allocation and behavior modeling