Skip to main content

The Socratic Trainer

The teacher of great teachers is the archetype you want leading your company's AI transformation. Socrates is the reference.

The Test That Sorts Trainers​

The most important fact about Socrates is that Plato had a student of his own named Aristotle, and Aristotle was as consequential a teacher as either of them.

Three generations of world-reshaping thinkers, each trained by the previous one. That is what "good teacher" means when you look at the downstream effects. One generation of transmission is an anecdote. Three is a lineage. If you want to evaluate a teacher, ask who their students' students are.

The ideal AI trainer passes this test. They can teach the material, and they can teach others to teach the material. Their mentees become trainers, and their mentees' mentees become trainers. The teaching propagates through the generations on its own.

Why This Matters For AI Specifically​

The AI field moves faster than any curriculum can follow. A trainer who teaches a specific tool, workflow, or prompt template is training people in something that will be obsolete in months. Every course that treats itself as the final answer is dated before it ships.

The Socratic trainer teaches something that does not go obsolete: how to think about the tools, how to evaluate what is worth learning, how to build a personal system that compounds, and how to pass that capacity to someone else. When the tools change (and they will change this quarter and the next quarter and the one after), the trainee's foundation still holds, because the foundation operates at the meta level.

See Learn The Harness, Not The Wrapper for the adjacent principle on tool selection. The same logic applies to pedagogy. Teach the harness of thinking; let the wrappers come and go.

The Sauce Recipe Generator​

A useful metaphor. Most companies hiring an AI trainer think they are buying the sauce: the specific template, the specific Jarvis configuration, the specific workflow. Which looks like a good deal. Get the pre-built thing, copy it, ship it.

The sauce is a trap. In a field that updates quarterly, the sauce expires faster than the invoice clears. What you actually want is the master chef: the person who can generate new sauce recipes whenever the kitchen changes. The sauce recipe generator.

Human capital here is divine capital: the image-bearer who can receive new patterns and translate them into sauce for whichever kitchen they walk into. A template is a snapshot of a generator's output. The generator is the asset.

If your budget allows for it, hire the generator.

The Four-Question Interview​

When evaluating an AI trainer, ask questions that surface the lineage.

  1. Who are their students, by name? Specific people, specific outcomes. "I have trained hundreds" without names is a marketing claim. "I trained Alex, who now runs workshops for their company's 300-person team" is a lineage claim.
  2. Who are their students' students? The killer follow-up. If the trainer's mentees are not training others, you are hiring a terminal teacher. If the mentees are running their own cohorts, you are hiring a Socratic one.
  3. How much of their material is open? A trainer who publishes their approach openly is demonstrating confidence that the value is in their thinking. See Raise The Floor for why this pattern wins at the organizational level and Permissionless Knowledge for why it wins at the ecosystem level.
  4. Can they teach without their own tooling? If they can only work inside their specific stack, they are a sales engineer for a product. If they can walk into any kitchen and still deliver, they are a chef.

The ideal answer to all four is obvious. Most trainers fail at least two. Hire accordingly.

What This Looks Like Inside AAS​

The Applied AI Society treats this architecture as a design requirement.

The Agentic OS Trainer role explicitly exists to produce trainers who produce trainers. The Supersuit Up workshop is open source. The trainer's guide that teaches people to run the workshop is published. Chapter leaders run the workshop locally, which produces more trainers, who run the workshop locally, which produces more trainers. That is three generations and counting.

Nothing is held back as "the real secret." The real secret is the human who understands the system deeply enough to improvise whenever the system needs to change.

The Hiring Implication​

If you are a business owner deciding who to bring in to lead your AI transformation, the selection criteria are simpler than you have been told.

You are hiring a generator of capable people. Curriculum is a byproduct. If the trainer's best outcome is "they ran a workshop and your team got a little better," that is short-term value. If the trainer's best outcome is "your team now has two people who can run this workshop themselves, and they are training a third," that is compounding value.

See Hiring Applied AI Practitioners for the general hiring framework and Four Levels of Applied AI for Existing Businesses for where in the adoption curve this kind of trainer pays off hardest.

The multi-generational version costs more up front. It is also the only one that earns its price back on a time horizon that matters.

Further Reading​