Unscripted Small Business Podcast

AI Job Description SOP: Ask What AI Can Help With and What Stays Human

AI Job Description SOP: Ask What AI Can Help With and What Stays Human

This is a Small Business SOP: a free, do-this-today standard operating procedure pulled from a real operator on the Unscripted Small Business podcast. This one comes from Barry Moline of BJM Solutions, who was a CEO for 29 years and now speaks on leadership, teamwork, influence and AI and the future of leadership.

Most small businesses start with AI from the tool side: someone finds an app, shows it off, and the team wonders what it means for their job. Barry starts from the job side. Update each job description with the person who holds it, hand it to your AI, and ask what it can help with and what should stay human. This SOP is that routine, plus the two builds that made a room of industrial-laundry operators sit up.

Barry Moline quote: When we know more about each other, we work better together.

The Play: One Line
Update each job description with the employee, enter it into your AI, ask what tasks AI can help with and what should remain human, then build the helpful ones from the manuals you already have.

Start AI adoption from the job description, not the tool

I asked Barry how a CEO brings AI into a company without building resistance. He didn’t start with a platform. He started with the paperwork every business already has: “Take their job descriptions and with their employees who have those job descriptions and update them to make sure they’re accurate.” Then: “Enter that job description into your favorite AI and say, what tasks can AI help with? And what tasks should remain human?”

Two things make this work for a small team. The employee is in the room from the first step, so the list is about their real job, not a guess about it. And the second question is built in: the answer always includes the work that stays human. Barry ran it with a group that operates industrial laundries, across five kinds of workers: maintenance techs, safety technicians, supervisors, floor workers and drivers who visit customers.

He speaks on “how leaders stay in charge of AI,” and this is a good example of what that looks like in practice. If your team is already nervous about AI, pair this with the AI Adoption SOP. If you want to roll it out in small weekly steps, the 1% Rollout SOP fits right behind it.

The numbers behind the play

All four come from Barry Moline on the episode.

29 yrs
Barry’s time as a CEO
5
worker categories in the industrial-laundry group he ran it with
2
questions: what AI can help with, and what should remain human
2x / month
the cross-department AI show-and-tell he recommends

The 8-step AI job description routine

1
Start small
Pick one role and pull its job description
Start with the role that has the most paperwork, lookups or training in it. One role, one description. You are not rolling out AI to the company today. You are finding out what it could do for one job.
2
With the employee
Update it with the person who does the job
Sit down with the employee and go line by line. Descriptions drift: Barry notes they lose relevance after six months or a year, and “you have to add some tasks or take some tasks away.” His rule is simple: “Make sure the job description is up to date.” And it is not a big project: “That takes a little bit of time, but not a ton of time.” Doing it together is the point. The person sees their real job on the page before any tool shows up.
3
The prompt
Ask your AI the two questions
Paste the updated description into whichever AI tool your business already uses, and ask exactly what Barry asks: “What tasks can AI help with? And what tasks should remain human?” His promise about the output: “It’ll give you a great list.” Save the answer next to the description.
4
Sort it
Sort the list: the hands-on work stays human
Go through the answer with the employee and mark each task. Physical work stays with people. In the industrial laundries Barry worked with, the things AI couldn’t help with were “like manual labor”: “Somebody’s got to turn the screwdriver. Somebody’s got to replace the belt.” Circle the tasks where the employee says “that would really help.” Those are your first builds.
5
First build
Build from the documents you already have
The laundry group’s best idea used manuals they already owned. They could “upload all of its operations manuals into their AI” and then “ask AI to create a safety course for individual employees who are working on the floor,” with test questions. Before you upload anything, run the AI Paste Audit SOP so you know what you are handing over. And have a supervisor check the course against the manual before anyone takes it.
6
Translate
Translate it into the language people learn best in
For workers who are bilingual but stronger in another language, translate the course. Barry’s reason: “You really want people to understand it at their core in their native language.” The team still works together in one language: “So they would still talk about it as a team in English, but at least they could learn everything in Spanish.”
7
At the machine
Put the manuals in the tech’s pocket
Load the operating manuals for every machine so the maintenance tech searches the problem instead of paging through binders. In Barry’s example, “the manager or the maintenance tech could look on his phone while he’s at the machine and then figure that out.” The AI gets you to the right page faster. The fix is still the tech’s.
8
Share and repeat
Show it to the team, then do the next role
Share what each role found. Barry recommends getting “cross-departmental teams together, anybody who wants to, twice a month, and share how people are using AI.” Peer demos lower the pressure: “The stress level is lowered significantly and our minds are open to learning from each other.” When he walked the laundry group through it, “their mouths were dropping open, like, I never thought about that.” Then pick the next role and start again at step 1.

A job description worksheet you can copy today

One row per task from the updated description. The first two rows are from Barry’s industrial-laundry example. Add your own below them.

Role Task AI can help Stays human First build
Maintenance tech Fix a machine Search the operating manuals Turn the screwdriver, replace the belt Manuals searchable on the phone
Floor worker Work safely Safety course with test questions Talk it through as a team Course translated into Spanish
Your role        

Barry gives away short one-to-two-page guides on building teams and influence at barrymoline.com/resources. They pair well with step 8, when you get the team in a room.

Take the whole SOP with you

One printable page: the play, the 8 steps, the two-question prompt, a fill-in worksheet and a checklist. No email required.

📥 Download the AI Job Description SOP (PDF)

Listen & watch

This SOP is distilled from Barry Moline’s episode on the Unscripted Small Business podcast:

The episode also covers the other half of Barry’s story: after one bad hire he switched to hiring for get-along ability first and training the skills, and he says it turned his organization around. That part is in the recap.

More operator SOPs, every week.

Real operators, real playbooks: free SOPs you can run this week. Browse the rest at Small Business SOPs and subscribe to Unscripted Small Business so you never miss one.