Is the LLM a tool or a new form of intelligence itself?
Does framing AI as merely delivering pre-existing intelligence miss what's actually happening? This explores whether the model itself constitutes a fundamentally new intelligence-medium with distinct cultural effects.
McLuhan's "the medium is the message" reframes the question of what a technology does. The content delivered by a medium (a TV show, a printed book) is what attention focuses on, but the medium's effects on the culture come from its medium-properties — its rhythms, its infrastructural demands, its social organization — not from any specific content. Television's effect was television-as-format; the shows mattered less than the form.
The same logic applies to LLMs. The standard framing treats AI as a delivery mechanism for some pre-existing thing called intelligence: AI helps you draft, AI helps you research, AI helps you think. The prior thing — intelligence — is taken as the message; AI is just the channel. This framing misdescribes the situation. There is no prior intelligence sitting upstream waiting to be delivered. The intelligence that AI produces does not exist in the absence of the model. It is constituted in the model's generative process. The model is not delivering intelligence — it is the intelligence-form being introduced into the culture.
What the model-as-medium does is make intelligence liquid. Intelligence had been illiquid — bound to specific people, requiring time to acquire, transferable only through teaching. The model converts intelligence into a generative format that can be invoked anywhere, by anyone, on any topic, instantly. Like printing converted knowledge into mass-reproducible form and changed what knowledge could do, the model converts intelligence into generative form and changes what intelligence can do. The cultural effect comes from this conversion — from the new medium-properties — not from any specific output.
This is why Where does the value of AI output actually come from? follows: a medium has no content of its own to value; value emerges in use. And Does AI actually commodify expertise or tokenize it? follows because tokenization is the medium-property of the model. Naming the model as the message keeps attention on the right level of analysis.
Source: Tokenization of Intelligence
Related concepts in this collection
-
Does AI actually commodify expertise or tokenize it?
The standard framing treats AI output like mass-produced commodities, but does AI's contextual, mutable nature fit better with token economics than commodity theory?
the medium-property the model exhibits
-
Where does the value of AI output actually come from?
If AI-generated intelligence has no intrinsic content-value like physical goods do, what determines whether it's valuable to someone? This explores whether value lives in the token or the receiver.
the value-theoretic consequence of model-as-medium
-
How do science fiction narratives about AI shape actual AI development?
This explores whether imaginaries of AI in fiction—from Čapek's robots to Singularity scenarios—function as self-fulfilling prophecies that causally influence the systems researchers build, creating a feedback loop between narrative and technology.
adjacent claim about how the model operates as a cultural object beyond its outputs
Click a node to walk · click center to open · click Open full network for a force-directed map
Original note title
the model is the message — the LLM is the intelligence-medium not a tool that delivers intelligence