Courses, AI, and the End of Trading Time for Money

"AI couldn't ever defeat you. AI is only as good as you are." Jasmine Jonte, founder of Cre8tion, said this to Hannah Chapman on The Expansive CEO podcast — and it reframes the anxiety a lot of experts feel about AI into something much more useful. Let's get into the conversation. 🎯

Q: Should experts be worried that AI is making their knowledge less valuable?

A: Jasmine's honest take: information was never actually the value — the client base just wasn't sophisticated enough yet to see that. Back around 2008-2015, curated information was genuinely valuable on its own, because finding a trusted, organized source beat piecing something together from scattered blog posts and forums. Now that anyone can get comprehensive information instantly from an LLM, the thing people are actually paying for has always been transformation, not information — it's just becoming obvious now.

Her reframe for anyone anxious about this shift: the very thing you're scared of — that your knowledge is "just information now" — is actually your advantage, because information curated in your specific voice, tone, and backed by your real experience is exactly what positions you to succeed in an AI-saturated world. The fear points directly at the unlock.

Q: Isn't there a risk that AI can just recreate an expert's knowledge on its own?

A: No — and this came from a compelling technical grounding in the conversation. Hannah's husband works directly in agentic AI development, and his core observation is one Jasmine agrees with completely: AI can only work with the information it's given. It's not extracting anything from a brain telepathically — it's predicting the next best output based on what's already been provided to it.

Jasmine sees this play out constantly with clients: when someone sends her an AI-generated course outline built without real instructional design input, it's consistently bad — bad enough that starting from scratch would often be faster. The built-in AI tools inside popular course platforms aren't much better. The pattern is consistent: AI without genuine expertise behind it produces garbage. AI trained on real expertise produces something genuinely powerful.

Q: What does it actually look like when AI is used well inside a course?

A: Jasmine's team builds custom GPTs, coach clones, and interactive dashboards trained specifically on a client's own body of work — not general internet content. Her standout example: a custom AI agent built for Biddy Tarot (the world's leading tarot education company) that helps students interpret tarot card meanings on demand. Someone pulls a card, doesn't know what it means, asks the agent, and gets a real interpretation plus reflection prompts — with a natural upsell into other programs woven in.

🎯 Why this particular tool works so well: it has repetitive, ongoing use. Jasmine contrasts this with something like a one-time bio generator — genuinely useful, but only used once a year. Tools built around a recurring need get used constantly, which means students stay deeply engaged with (and dependent on) the creator's specific brand. As Jasmine puts it: "I can't leave [this creator] because I've got to keep access to that custom GPT."

Q: What's the risk of relying entirely on AI to build a course from scratch?

A: Garbage in, garbage out — literally. Both Jasmine and Hannah pushed back on the idea of fully outsourcing course creation to AI with no expert refinement. Jasmine's clients occasionally try prompting a course into existence entirely on their own, and the resulting outlines are consistently weak — not because AI is a bad tool, but because most people don't know instructional design well enough to give AI the right frameworks to work from.

⚠️ The practical takeaway: treat AI as a powerful thought partner and drafting tool — genuinely useful for generating a first pass, formatting, or repurposing content into different formats — but not as a replacement for real instructional design or your own lived expertise.

Q: Does the rise of AI make human connection more or less important in course design?

A: More — significantly more, according to Jasmine. Her reasoning: a few years ago, simply getting information out into a course format took so much time and energy that there was little left over to focus on real connection. Now that AI has made producing the informational content faster, that freed-up time and energy can go toward the human experience surrounding it. She sees this as a genuine net positive, not a threat.

This shows up concretely in what Jasmine calls support structures — everything wrapped around the core curriculum. The old standard (a generic weekly Q&A call) doesn't cut it anymore; students today are more sophisticated and want to understand the actual purpose and facilitation plan behind every touchpoint they're offered. Meanwhile, more creators are building hybrid models — blending group programs, occasional 1:1 touchpoints, and even in-person retreats — because that layered structure amplifies transformation in a way pure content alone can't.

Q: How do you decide the right amount of support to build into a program?

A: It has to align with your specific ideal client's actual life — not a generic assumption about what support "should" look like. Jasmine shared a personal example: she once tried launching a multi-day retreat that flopped, and market research afterward revealed the real issue — her audience genuinely couldn't commit to being away from their kids for that long. Meanwhile, busy seven-figure entrepreneurs in other programs often prefer async support (sending a question and getting a Loom response back) over live calls entirely.

🎯 Her framing for getting this right: aim for the "Goldilocks amount" — not too much, not too little support. This matters on the sales side too: piling on excessive support (weekly two-hour calls, multiple retreats, an overloaded community) can actually hurt conversion, because it makes the commitment feel overwhelming rather than reassuring.

Q: Why do some creators over-deliver content or support to the point of burning themselves out?

A: Jasmine connects this directly to confidence, not generosity. She's observed a consistent pattern: creators overload their programs — a "500-course library," or five modules that are each an hour long, or a 300-page resource dump — as a way of trying to prove their own value, often rooted in a deeper "not enough" feeling rather than genuine student need. Nobody actually wants that volume; it just feels safer to give than to trust that a tight, well-designed program is sufficient.

 The same dynamic can show up on the support side: creators who don't feel confident in their curriculum sometimes over-invest in support (extra calls, extra retreats, an overbuilt community) to compensate, rather than addressing the root issue — a program that isn't as strong or clear as it could be.

Q: Pricing seems to be a recurring struggle for creators. What patterns do you see?

A: A near-universal one: most clients arrive unsure what they can charge, and consistently underestimate it. Jasmine described a common arc — someone comes in planning to charge $5,000, and once the program is actually built with real expert input, they realize the value is so much greater than they assumed that they end up doubling the price, because they now genuinely believe in what they built, having previously carried the entire weight of building it themselves without outside validation.

Her team's marketing lead often gives clients direct, confident pricing guidance — and Jasmine's observed that clients trust that external confirmation in a way they can't quite give themselves. Sometimes the missing ingredient isn't a better pricing formula — it's someone else believing in the program enough to say "yes, that's the right price," out loud.

The Bottom Line

AI doesn't threaten real expertise — it exposes the gap between creators who have genuine, curated experience to offer and those who were only ever selling raw information. Build AI tools around your recurring, repeatable knowledge rather than trying to replace your own instructional judgment with a prompt. And as content creation gets faster, invest the time you save into genuinely thoughtful support structures — sized to what your specific audience can actually commit to, not what feels safest to overprovide.

Curious what a genuinely well-trained AI tool could add to your own program? 🚀 Head to cre8tion.co to talk through what that could look like.

Enjoyed this conversation? Subscribe to The Expansive CEO for more real conversations on business, money, and building a life that's genuinely expansive.