Every office has one: the AI Joker. They use AI every day, send emails they never read and trust numbers they never checked. After the third close call, a leader starts asking the obvious question: should we just replace them?
It feels decisive. But in most cases, replacing someone for poor AI habits is the more expensive, slower and riskier option. Here’s why, and how to decide.
The hidden costs of replacing
The cost of replacing an employee is much more than a recruiter’s fee. Leaders often underestimate:
- The empty seat. Work slows or stops while the role is open, and the rest of the team picks up the slack.
- Hiring and onboarding. Job ads, interviews, management time, paperwork and weeks of getting someone new up to speed.
- Lost knowledge. Your current employee knows your clients, your systems and how things really get done. That walks out the door with them.
- Team morale. When someone is let go for “not getting AI,” everyone else wonders if they’re next, and starts hiding how they use it.
- The same problem again. Unless the new hire was trained somewhere else, there’s a good chance they have exactly the same habits. You’ve paid to replace one AI Joker with another.
Poor AI habits are rarely a people problem. They’re a training problem. Nobody showed them how.
What upskilling keeps, and what it adds
Upskilling keeps everything that makes an employee valuable (experience, relationships, business knowledge) and adds the one thing that’s missing: the right AI skills, tools and rules.
- Faster results. New skills are applied to real work straight away, instead of waiting months for a new hire to settle in.
- Loyalty. People remember when a company invests in them.
- A ripple effect. One trained AI Ace often becomes the person colleagues ask for help, spreading good habits across the team.
- Lower risk. Training, approved tools and clear rules reduce the data and quality risks that caused the problem in the first place.
Replace vs upskill at a glance
- Cost: replacing means recruitment, onboarding and lost productivity; upskilling is a focused investment in someone you already employ.
- Speed: a new hire takes months to reach full speed; upskilled employees apply new skills from the first sessions.
- Knowledge: replacing loses it; upskilling keeps it and builds on it.
- Risk: a new hire may bring the same habits; upskilling comes with rules, approved tools and SOPs.
When replacing might still be right
Upskilling isn’t a magic fix for everything. If someone has been trained, given clear rules and the right tools, and still refuses to follow them, that’s a conduct issue, not a skills gap. But that should be the conclusion after training, not the assumption before it.
How to upskill an AI Joker in four steps
- Assess. Find out what they know, what they don’t, and which habits put the business at risk.
- Train on real work. Hands-on training using their actual weekly tasks, not generic theory.
- Set the rules. Approved tools, data rules and a simple AI policy they can follow.
- Make it stick. Step-by-step SOPs for the tasks where AI helps most, plus hands-on support until the new habits are second nature.
The bottom line
Before you replace your AI Joker, ask whether anyone ever trained them. In most cases, the cheaper, faster and kinder option is also the smarter one: don’t replace them, upskill them.
Upskill one team member or the whole team.
Start with an AI skills assessment. It shows exactly what each person needs, and what it will take to turn your AI Jokers into AI Aces.
Get My Team Assessed ↗