AI as thinking partner, not replacement.
Extended Cognition & AI
The Core Idea
This is the thesis behind Prosody Learning and most of what I build with AI. It comes from Clark and Chalmers' Extended Mind thesis — the idea that cognitive processes don't stop at the skull. When you use a notebook to remember, the notebook is part of your cognitive system. AI has the potential to be the most powerful cognitive extension humans have ever had — but only if it's designed to scaffold thinking rather than substitute for it. The same tool that helps you think deeper can also make you think less. The design choice is everything.
Components
Scaffold vs. Substitute
The same AI capability can either extend or atrophy human cognition depending on how it's designed. An AI that gives you the answer makes you weaker over time. An AI that helps you construct the answer yourself makes you stronger. This is the most important design decision in AI-assisted learning.
ChatGPT giving a sales rep the perfect discovery question to ask is substitution — the rep never learns to formulate questions themselves. Prosody asking 'What do you think the buyer is really worried about, and what question would surface that?' is scaffolding.
Cognitive Mirror
AI that reflects your own thinking back to you in a structured way, making implicit reasoning explicit. This is powerful because most people can't see the gaps in their own logic until someone (or something) maps it out for them.
After a rep debriefs a call, the AI doesn't say 'here's what you should have done.' It says 'You identified budget and timeline but didn't explore decision process. What do you think happens when you advance a deal without understanding how they buy?' The rep sees their own gap.
Thought-Action Gap
The distance between knowing what to do and actually doing it. This is the gap that kills most training ROI. Reps know they should ask about business impact. In the moment, under pressure, they revert to feature pitching. AI that operates in the workflow — not just the classroom — can close this gap.
Real-time AI coaching that notices a rep has spent 8 minutes on product demo without asking a single question about business outcomes, and prompts them mid-call with a nudge — not a script, a nudge — to reconnect to the buyer's priorities.
Complementary Cognition
Humans and AI each contributing what they do best. AI excels at pattern recognition across large datasets, consistent application of frameworks, and memory. Humans excel at empathy, creative problem-solving, and reading emotional subtext. The best systems leverage both.
AI analyzes 500 closed-lost deals and identifies that deals stall 73% of the time when economic buyer engagement drops after Stage 3. The human coach takes that insight and designs a conversation with the rep about why they're avoiding the executive conversation — which turns out to be a confidence issue, not a skill issue.
In Practice
Prosody Learning is built entirely on this framework. When a sales rep practices a discovery call in Prosody, the AI never tells them what to say. It plays the role of a realistic buyer and then, in the debrief, asks Socratic questions that help the rep see their own patterns. 'You asked three closed-ended questions in a row — what do you think that did to the buyer\'s willingness to share?' The rep constructs the insight themselves, which means it actually encodes in long-term memory. Early testing showed reps who used Prosody's scaffolded approach retained methodology 3x longer than those who just reviewed call recordings with AI-generated summaries.