When anyone can ask a chatbot, what are paid courses still for?

There's a question that keeps turning up in the communities we follow, and it isn't "which LMS should I pick".

It's whether making courses for a living still makes sense when anyone can ask a chatbot. Creators in Europe are planning Linux and software-engineering courses anyway, plus mini tutorials that funnel into full programmes, and getting replies that courses still sell if they solve real problems. First-time eLearning founders who built voice-simulation practice for deskless field staff — after lecture-and-booklet training failed — are asking what running an eLearning company actually means: proof, packaging, pilots. Creators who left marketplaces to own checkout are validating payment providers and then hitting LMS edge cases. Instructional designers and small-business owners are shopping LMS stacks for customer training, partner enablement, continuing education or selling courses — and they are not shopping for "AI course generators". Builders keep arguing that generic "intro to X" dumps are dead, while guided labs, failure modes and checked exercises still sell.

We build tools around learning and membership, so we have a bias toward people who teach. We also think this debate is unsettled enough that arguing a side would be dishonest. Below is the framing we keep hearing, then the questions we can't shake.

The "chatbots killed courses" side

The case is straightforward. If a learner can paste a stuck homework question into a chatbot and get a working answer in seconds, why pay for a recorded lecture that covers the same ground more slowly? If a chatbot can outline a curriculum, draft quizzes and summarise a textbook chapter, what is the course creator adding?

People on this side tend to point at:

• Commodity topics that are already saturated with free explainers
• Courses that are mostly slideshow voiceovers with no practice
• Catalogues that promise "become an X in 30 days" without a checked outcome
• Buyers who churn after realising the chatbot in their browser does half the job

The uncomfortable version of this argument is not "AI is smarter than teachers". It is "a lot of paid courses were already thin, and chatbots made the thinness obvious".

The "chatbots killed the generic ones" side

The pushback is just as consistent. Courses still sell when they do something a chatbot is bad at alone: holding someone through a sequence, checking their work against a standard, putting them in a lab where failure is safe, connecting them to peers who have the same job, or certifying a skill an employer or regulator recognises.

People on this side tend to point at:

• Guided labs and failure-mode walkthroughs, not topic dumps
• Practice that needs a human scenario (the voice-simulation example for field staff sits here)
• Cohorts, deadlines and accountability that a solo chat session doesn't provide
• Customer and partner training where the company cares that people finished and can do the task
• Continuing education where a score or certificate has to mean something outside the course

The free-content funnel sits in the middle: give away the explainers (or accept that chatbots will), and charge for the labs, feedback and proof.

The grey area we keep circling

Plenty of creators are already living in the middle without calling it a philosophy. Mini tutorials and public write-ups pull people in. The paid course is where practice gets checked. Marketplace leavers who now own Stripe or Paystack checkout discover that the hard part after payment is renewals, quiz retakes and certificates — not the sales page. That is a product problem disguised as a marketing one, and chatbots don't remove it.

All-in-one platforms versus assembled stacks show up in the background of these talks too. We are not reopening that whole debate here. The useful overlap is simpler: whichever stack you use, the buyer is paying for an outcome, not for a video folder.

Questions we can't shake

What still justifies the price?

• If a chatbot can answer the same questions, what is the learner actually buying? Feedback? Sequence? Community? A credential? Access to the instructor?
• Which parts of your course would still be worth paying for if every lecture transcript were free tomorrow?
• For employers and partners, is "completed the course" enough, or do they need evidence the person can do the job?

What did chatbots actually kill?

• Are generic "intro to X" courses dying where you are, or is that mostly loud anecdotes?
• Have you seen a course lose sales specifically because buyers said they would just ask a chatbot instead?
• Conversely: have chatbots helped your course business — as tutors inside lessons, as draft assistants for you, or as a reason buyers value human feedback more?

What should replace the thin course?

• If you were building a paid offer from scratch now, would you start with a cohort, a self-paced lab, a simulation, a certificate track, or something else?
• How do you check exercises at a price people will pay? Peer review, auto-graders, instructor hours, scenario scores?
• Where does free content stop helping the funnel and start cannibalising the paid offer?

What does "running an eLearning company" mean once the novelty fades?

• For first-time founders: what counts as proof that a pilot worked — completion, on-the-job change, a client renewal, something else?
• After checkout works, which edge case ate your week: renewals, refunds, retakes, certificates, access fixes?
• Are you selling to individuals, to teams, or to partners — and did that choice change what the course had to prove?

Where we are, honestly

Our own lean: chatbots raised the floor for free explanation and lowered the value of courses that were mostly explanation. They did not remove the need for practice, feedback, accountability or proof. That is easy for us to say from the tooling side. It is harder to live when you are pricing a first course in a market where every buyer has a chatbot open in another tab.

So we'd rather hear from the people selling, buying and assigning these courses.

Your turn

1. If you sell or assign courses, what still makes someone pay after they already have a chatbot — and what stopped selling?
2. Builders and instructors: which format are you betting on now — cohort, lab, simulation, certificate track, or something thinner — and why?
3. First-time eLearning founders: what did "the pilot worked" look like in practice, not on the sales page?
4. Have chatbots changed how you design practice and assessment, or only how you write marketing copy?

Tell us what still sells where you are. We are especially interested in the unglamorous offers that keep renewing, not only the ones that launch well.