← Protocols for Business

September 20, 2026

SIG-P4B 2026-09-20

Participants: rafa (UTC+1), Sachin

This session was primarily a working/planning meeting between rafa and Sachin ahead of a workshop they are co-facilitating the next day. It opened with a conceptual discussion refining the definition of "kits" and the role of protocolization in the AI environment, then shifted into practical preparation: facilitation roles, agenda timing, attendee logistics, and the state of the shared GitHub/Discord setup and "context tank." Much of the second half is operational detail about registrations, calendar invites, and exercise sequencing.

  • **Sachin — refining the "kit" definition via the car analogy:** He revisited the comparison between AI kit-building and early car tinkering. He argued that early cars existed in a "very low protocolization environment" (no roads, highways, traffic laws, or authorities), whereas AI kit work happens in a "much more heavily prototyped/protocolized world."
  • **Sachin — a better car analogy:** He suggested Ballard's *Crash* and *Concrete Island* as better comparisons, since those depict figuring something out within an environment that is already heavily built up, rather than starting from scratch.
  • **Sachin — evidence for protocolization framing:** He grounded the claim in a Protocol Institute graph showing rising usage of the word "protocol" after the Second World War.
  • **Sachin — distinguishing protocolization from skill environments:** Early car tinkerers already had machine-tinkering skills; that reflects an environment where relevant skills existed, which he distinguished from true protocolization. By contrast, AI sits atop many existing protocols (e.g., Git/GitHub, various frameworks).
  • **rafa — is protocolization relative/perceptual?** He pushed back by asking whether we only see our environment as "highly protocolized" from our vantage point — would a farmer of that era have felt they were in a protocolized environment? He speculated we might actually be in a comparatively *non*-protocolized environment we can't yet perceive, and that in fifty years everything may look "hyperprotocolized" by comparison (analogous to a farmer being unable to conceptualize farm automation).
  • **rafa — AI as "oozy":** He noted AI is special because it is "oozy" and itself protocolizes environments in some ways, suggesting the dynamics differ from the car case.

Session Recording Summary · 38m 13s · Full notes ↗

Reading: Not clearly identified in the discussion. The conversation centers on the participants' own workshop materials — an "AI CMM" (AI Capability Maturity Model) slide deck and a "kit" framework — rather than a shared external reading. The J.G. Ballard novels *Crash* and *Concrete Island* are referenced as conceptual comparisons but were not the session's reading.

This session was primarily a working/planning meeting between rafa and Sachin ahead of a workshop they are co-facilitating the next day. It opened with a conceptual discussion refining the definition of "kits" and the role of protocolization in the AI environment, then shifted into practical preparation: facilitation roles, agenda timing, attendee logistics, and the state of the shared GitHub/Discord setup and "context tank." Much of the second half is operational detail about registrations, calendar invites, and exercise sequencing.

  • **Sachin — refining the "kit" definition via the car analogy:** He revisited the comparison between AI kit-building and early car tinkering. He argued that early cars existed in a "very low protocolization environment" (no roads, highways, traffic laws, or authorities), whereas AI kit work happens in a "much more heavily prototyped/protocolized world."
  • **Sachin — a better car analogy:** He suggested Ballard's *Crash* and *Concrete Island* as better comparisons, since those depict figuring something out within an environment that is already heavily built up, rather than starting from scratch.
  • **Sachin — evidence for protocolization framing:** He grounded the claim in a Protocol Institute graph showing rising usage of the word "protocol" after the Second World War.
  • **Sachin — distinguishing protocolization from skill environments:** Early car tinkerers already had machine-tinkering skills; that reflects an environment where relevant skills existed, which he distinguished from true protocolization. By contrast, AI sits atop many existing protocols (e.g., Git/GitHub, various frameworks).
  • **rafa — is protocolization relative/perceptual?** He pushed back by asking whether we only see our environment as "highly protocolized" from our vantage point — would a farmer of that era have felt they were in a protocolized environment? He speculated we might actually be in a comparatively *non*-protocolized environment we can't yet perceive, and that in fifty years everything may look "hyperprotocolized" by comparison (analogous to a farmer being unable to conceptualize farm automation).
  • **rafa — AI as "oozy":** He noted AI is special because it is "oozy" and itself protocolizes environments in some ways, suggesting the dynamics differ from the car case.

Questions & Disagreements: - **Is protocolization objective or perceptual?** The main friendly disagreement: Sachin leaned on the Protocol Institute graph as a rubric for increasing protocolization over time, while rafa questioned whether "protocolized" is a relative judgment that we can't fully assess from inside our own moment. - **Whether to include failure modes / success patterns (AI CMM slide 25) in tomorrow's worksho

Participants: rafa (UTC+1), Sachin