Training for AI Roles That Didn't Exist Yet
When the roles are brand new, off-the-shelf training cannot exist yet, and that changes who can fill them.
You cannot buy a course for a job that did not exist last year.
That single fact is the whole problem with training for new AI roles. Training for a new AI role means building curriculum from the expertise your own team is inventing right now, because no vendor has had time to write a course for a job that is only months old. The big content libraries are built by recording what is already known and packaging it, so by definition they trail the frontier. For exactly the roles companies most need to fill, the off-the-shelf catalog is either empty or already out of date.
Rewind to 2022. Before that year, almost no job posting asked for "prompt engineering," "AI workflow design," or "agent operations." Today, AI and big data top the World Economic Forum's list of the fastest-growing skills employers want (Future of Jobs Report 2025). Companies are scrambling to fill positions that were invented, in some cases, last quarter.
Why AI layoffs are really a training problem
Here is the part the layoff panic misses. The clearest documented case is customer service, where Gartner predicts that by 2027, half of the companies that cut staff citing AI will rehire for similar roles, often under new job titles (Gartner, February 2026). Pair that with the World Economic Forum's finding that 39 percent of workers' core skills will be transformed or become outdated between 2025 and 2030 (Future of Jobs Report 2025), and the shape of the trap is clear.
Companies are letting people go on the theory that the future needs a different workforce. But the future roles are so new that there is no proven pipeline to hire for them and no library to train for them. So the same organizations end up bidding against each other for a tiny pool of people who already have the new skill, while the institutional knowledge they just shed walks out the door. They optimized for a skill they could have built and discarded the context they cannot buy.
Where does training for a new AI role come from?
If no vendor has the course yet, the knowledge can only come from inside your own walls. Somewhere in your company, a handful of people have figured out how to use these tools on your actual work: your data, your customers, your processes. That hard-won, undocumented expertise is the raw material for the training nobody can sell you.
The catch has always been turning that scattered expertise into something teachable before it goes stale. Traditional course production takes months. By the time the course on the new AI workflow ships, the workflow has changed, and you are teaching last season's version.
How to build training for new AI roles in minutes
This is the specific constraint Honen removes. You hand it the recordings of your best people doing the new work, the rough notes, the internal docs, or just a topic, and it builds a real, structured course in minutes, with hands-on practice and a tutor that teaches from your material. On one recent 30-minute partner call, Ryan Trattner, co-founder and CTO of StudyFetch (Honen's parent company), built several complete courses before the call was over. When the tool or the process changes, you tell it what changed and it refreshes the affected lessons without resetting anyone's progress. You are turning your own frontier into curriculum as fast as you reach it.
What this means for your team
The hardest roles to fill right now are the ones with no curriculum, and that is exactly why they reward companies that can write their own. Off-the-shelf training will always trail the new work, because it can only package what is already settled. The organizations that win the AI transition will be the ones that take the expertise their own people are inventing this quarter and turn it into training the rest of the workforce can learn from next quarter. The same logic runs through reskilling instead of firing and doing reskilling for real.
AI moved the most valuable training inside your own building, into the heads of the people doing the new work for the first time. The question is whether you can get it out of their heads and into everyone else's before the role changes again.
Frequently asked questions
Why can't you buy training for new AI roles?
Course libraries are built by packaging knowledge that is already settled, so they trail the frontier by months or years. For roles that are only a quarter or two old, no vendor has had time to write the course, which means the training has to be built from your own team's expertise.
What is the difference between upskilling and training for a new AI role?
Upskilling makes someone better at the job they already have. Training for a new AI role builds a capability that did not exist in the company before, usually from the undocumented expertise of the few people already doing the work.
How fast can you build a course for a new AI role?
In minutes. On one 30-minute partner call, our co-founder and CTO built several complete courses from existing materials, far faster than the months a traditional production cycle takes.
How do you keep AI training current when the tools keep changing?
When a tool or process changes, you update the source material and Honen refreshes the affected lessons without resetting learner progress, so the course keeps pace with the work.
Have a brand-new role and no course for it yet?
By StudyFetch Staff. Honen turns the materials your team already has into real courses they'll actually finish, built in minutes and measured by mastery.
