Why a Workshop, Not a Webinar

Webinars broadcast. Workshops build. Both are useful, and neither one substitutes for the other. When we tell people the IamHITL room is a workshop, we mean something specific about the shape of the hour, the position of the participant, and what leaves the room at the end.

This is a naming question with consequences. Pay for a workshop, get a webinar, and you leave with notes about someone else's practice. Pay for a webinar, get a workshop, and you leave with an artifact you did not expect to have to build. The label needs to match the room (Class C).

The pedagogical difference, in one paragraph

A webinar is a one-to-many broadcast. The speaker holds the map, the audience receives it, and success is measured in what was transmitted. A workshop is a small-group build. The facilitator holds a method, the participants hold the work, and success is measured in what each participant made and can still run without the facilitator in the room. In active-inference language: a webinar updates the audience's model from the speaker's, a workshop updates each participant's model from their own action and observation (Class E, drawing on the standard active-inference literature summarized in Parr, Pezzulo, and Friston, 2022).

Neither is bad. They are not the same product.

The market blurs the two labels. A recorded talk with a Q&A window at the end gets sold as a workshop. A working session with three participants and a facilitator gets sold as a webinar. Blurring is not a scandal, it is a habit of the industry, but it costs the buyer, because the buyer sets expectations from the word. Our position is narrow: for the outcomes IamHITL is trying to produce, the room has to be a workshop, and the label has to say so. That is not a claim that webinars are worse in general. It is a claim about what a specific outcome requires (Class C).

Four honest tests you can run at the door

If a session is a workshop, four things are true. Any of them missing is a falsifier that we invite (Class F).

Cohort size is small enough to see every participant. A room where the facilitator cannot name each person's current move is a room where the facilitator is broadcasting, whether or not there is a slide deck.

The participant does the work in the room. Not homework after, not a worksheet for later. The keyboard opens, the notebook opens, and the participant makes the thing while the facilitator watches and adjusts.

The artifact stays with the participant. In a form they can open on their own machine, without our tooling. If the artifact only exists inside our platform, we sold access, not a build.

The loop is visible. Plan, act, observe, repair, spoken out loud, run more than once. If nobody names the loop and nobody repairs anything, the session was closer to a demo than a build (Class B).

None of these tests is exotic. A boring test is the kind you can run without our help. If a session we run fails any of them, tell us.

Why we do not run webinars as our main product

We could. The economics are attractive: one speaker, many attendees, a recording that sells for a long time. Other people run that model well. We do not run it as our main product because the outcome we care about, a participant who can run the loop after we leave, does not reliably come out of a broadcast. It comes out of a room where the participant did the loop with a partner-in-practice sitting next to them (Class C). A recording is a document of what happened, useful for review; it is not, on its own, a workshop for the person watching it later (Class F).

Where the uncertainty lives

We do not have a controlled study comparing outcomes from webinars and workshops for the specific work IamHITL does. What we have is the recurring pattern in our own sessions and the standard finding in the active-inference and adult-learning literatures that participation, observation, and repair change the participant's model more reliably than reception alone (Class E for the literature, Class U for any strong causal claim specific to our format). If you have run the comparison and have data that cuts the other way, we want it.

A note on the wider program

Universal Natural Intelligence is a working hypothesis on an attainable path toward General Natural Intelligence: a natural, active-inference approach whose evidence is growing, evidence-classed, and tested in the open. Do not take the claim on faith. Test the build, inspect the gates, and help us find where it fails. The workshop is where that hypothesis meets someone else's hands (Class C).