Stop asking human or agent. Ask who becomes your next leader.
A legal marketing CEO in Dallas, a founder in Brisbane and a retailer in Tallinn on what AI should take, what people keep, and who gets to learn the job.
Automate the junior work without a plan, warns a Dallas legal marketing CEO, and "you don’t save money-you create a future leadership gap."
A Brisbane founder adds the twist: refusing AI is just as risky.
Meanwhile, a Tallinn retailer has a simpler test. What does a mistake cost?
Three leaders, three countries, and one pressure point: who does the entry-level work now.
Cheap mistakes go to the machine.
Jean-Christophe Thery, Founder of MusaArtGallery (Tallinn, Estonia)
“If mistakes hit margins or reputation (international shipping claims, chargebacks, refunds), we keep humans in the loop.”
“If a task is highly repeatable with low error cost, we automate.”
The rule is a price tag on being wrong. It is the least sentimental test in the piece, and the easiest to apply on a Monday morning.
AI drafts. Juniors still need the reps.
Nicole Farber, CEO of ENX2 Legal Marketing (Dallas, United States)
“In legal marketing, AI can take the first pass on repetitive work: content outlines, keyword research, reporting summaries, social post drafts, and campaign checklists.”
“Juniors still need reps, so let them QA AI work, join strategy calls, write post-project lessons, and learn why decisions were made. Otherwise, you don’t save money-you create a future leadership gap.”
Look at what she keeps: not the first pass, but the learning around it. Juniors move from doing the draft to checking it, and she wants them in the strategy room while they do.
Pick your risk: automate everything or refuse AI.
Rachael Karlyl, Founder & Chief Motivational Officer of SPRKL (Brisbane, Australia)
“Automate everything and you risk dismantling the development pathways that produce future leaders. Refuse AI and you risk becoming stale, uncompetitive and unattractive to ambitious talent.”
“Cutting headcount without considering the consequences isn't a workforce strategy.”
She objects to both extremes, and she is not choosing a camp. One risk is a missing pipeline of leaders, the other is a company ambitious talent walks past.
Where they split
Where AI belongs: the Tallinn retailer and the Dallas CEO agree. Repeatable, low-risk work goes to AI, and the first pass is the obvious place to start.
The junior question: Dallas and Brisbane both see a development problem if entry-level work disappears.
The common thread: nobody argues for zero AI, and nobody argues for zero people.
My take
These leaders agree that repeatable, easy-to-check work can go to AI. The catch is that this is also where juniors used to learn. A headcount plan that ignores that is a plan with a hole in it, and the hole shows up in your leadership team later. Before you sign off on the next automation, ask what the people who used to do that work were learning from it.
Your turn
The Brisbane founder says we need to "stop developing tomorrow's leaders for yesterday's jobs". Now that AI takes the first pass, where do your juniors get their reps?
Next: one AI founder reports agents automating 95% of billing tasks, a CPG trade-spend CEO limits automation to flagging, and an advisor watched founders hit walls automating pricing and hiring.
