Let AI do the admin so you can do the relationships
AI is brilliant at busywork and poor at feeling. Hand it the first, keep the second, and your network gets the time it deserves.
Listen to the episode
How Leaders Can Stop Getting Left Behind by AI
Building Your Leader Brand, Bob Gentle with Martin Wheatley
Most of the talk about AI is about what it will take away. In this episode, Bob and Martin Wheatley talk about what it gives back. Used well, AI strips out the busywork that eats your week and frees you up for the work only you can do.
For an expert business, a big part of the work only you can do is relationships. The thoughtful message, the well-timed introduction, the call that turns a past client into a referral. AI can't do those for you. It can clear the space for them.
AI is good at systems, people are good at feelings
A point Bob makes in the conversation is worth holding on to. AI is excellent at structure, process and systems. It struggles with feeling and emotion, with sensing how the person on the other end will react. That's exactly the part of a relationship that matters most.
The sensible split is simple. Let AI handle the parts of staying in touch that are really admin: working out who's due, pulling up what you discussed last time, drafting a first version, logging what happened. Keep the parts that need judgement and warmth for yourself.
Treat AI like someone you delegate to
The episode makes the case that the most useful AI skill isn't technical at all. It's delegation. Choosing the right job, explaining what good looks like, checking the result, and correcting it if it's off. Anyone who has managed people already knows how to do this.
The easiest way to start, the conversation suggests, is with something you actually need to get done. Not a course, not an experiment for its own sake. A real task you'd like to take less time.
Staying in touch is an ideal first candidate. It's a recurring job, it follows clear rules, and everyone knows they should do more of it.
In SignalBox: connect Claude or another AI assistant through SignalBox's MCP server. Then you can delegate the admin in plain English:
- "Who's due a follow-up today, and what did I last talk to each of them about?"
- "Draft a short check-in for each person due in my Clients circle. Keep it warm, no selling."
- "Add a note to Sarah's contact: she's moving to a new role in January, follow up then."
- "Log that I called Tom and he's keen on a workshop in the spring."
Your assistant reads your contacts and due follow-ups from SignalBox, writes notes and logs actions back to it, and leaves you to read the drafts, change what doesn't sound like you, and send the ones that feel right.
Talk, don't type
Both Bob and Martin recommend dictation tools that turn speech into text. Typing makes you edit as you go. Speaking lets you get everything out, then let the AI tidy it up.
This is a natural fit after a call or a meeting. Instead of promising yourself you'll write notes later, talk for thirty seconds about what you discussed, what they care about and what you promised. Ask your assistant to add it to their contact in SignalBox. Next time they come due, the context is already there.
Don't automate the relationship itself
There's a warning in the episode too. Strip out everything human and you end up average, because AI produces the most plausible average answer. A business that automates every message will sound like every other business that did the same.
The goal isn't to have AI send your follow-ups. It's to make sure you never skip one because you ran out of time or forgot. The judgement about what to say, and whether to say it, stays with you.
Try this week
- Connect your AI assistant to SignalBox and ask it who's due a follow-up today.
- After your next call, dictate your notes and have them added to the contact.
- Ask for draft check-ins for one circle, then rewrite each one in your own words before sending.
The leaders who get the most from AI won't be the ones who automate the most. They'll be the ones who use the time it frees to show up for people.