Where AI Actually Saves Time (Hint: It’s Not Where You Think)
AI and workflow is a topic that gets flattened into “efficiency” far too quickly.
The work isn’t slow because communicators are inefficient. It’s slow because the work is layered, contextual, and full of invisible steps that don’t show up on a project plan.
This edition isn’t about AI “doing the work” for communicators. It’s about where it actually saves time and energy in a real week, not by automating judgment or creativity, but by reducing the friction around the work. The setup. The sorting. The reformatting. The mental bookkeeping. The parts that drain attention before the real thinking even begins.
Where time is actually saved
For me, the biggest shift wasn’t adding new tools. It was letting AI sit inside the workflow instead of treating it like a one‑off experiment.
That doesn’t mean handing work over to it. It means using it where the work is mechanical, repetitive, or cognitively noisy and keeping the parts that require judgment firmly human.
Here’s where that shows up most clearly.
1. Raw input → orientation → sensemaking
Most weeks start the same way: meetings, interviews, half‑formed requests, and decisions that change mid‑sentence.
This is where AI saves me time reliably.
Generating meeting note summaries so I don’t have to reread full transcripts
Producing interview transcripts so I can focus on listening and follow‑ups in the moment
Pulling out decisions made, open questions, and next steps
What it gives me isn’t a draft — it’s orientation. A faster way to understand what just happened so I can decide what matters.
What stays human: interpreting nuance, sensing tension, understanding what wasn’t said.
2. Drafting → comparison → stress testing (not replacement)
I’ll be honest: I’ve experimented with AI first drafts. The results are usually bad — thin, generic, or missing the organizational reality entirely. Sometimes they’re useful. Often they’re not.
Where it does help is as a comparison tool.
I’ll write my own first draft, then ask AI to do one
Or I’ll ask it to stress‑test clarity: “What questions would frontline staff still have?”
Or I’ll use it to surface where a message could be misread, oversimplified, or challenged
This isn’t outsourcing writing. It’s pressure‑testing thinking.
What stays human: authorship, narrative decisions, accountability for the words on the page.
3. One message → many channels → less rework
This is one of the most practical time‑savers.
Once a message is solid, AI helps with the grind:
Reformatting for email, Teams, intranet, or leadership talking points
Adjusting length and cadence without changing intent
Maintaining consistency while avoiding copy‑paste errors
I still review everything. But I’m no longer rewriting the same idea five different ways from scratch.
What stays human: knowing which channel matters most, when not to post, and what tone the moment calls for.
4. Weekly recap → reflection → continuity
This is an under‑discussed use — and one I rely on.
At the end of the week, I use AI to help draft a weekly recap:
What went out
What shifted
What decisions were made
What’s carrying forward
I edit, add context, and correct tone — but the recap exists, which is the point. Without help, reflection is often the first thing to fall off the list.
What stays human: interpreting patterns, deciding what to adjust next week, and naming what matters.
What this isn’t
This isn’t about speed for speed’s sake. It isn’t about replacing communicators. And it definitely isn’t about pretending AI “thinks” for you.
Most of what I use AI for falls into three categories:
Reducing friction
Stress‑testing thinking
Maintaining continuity
Some experiments stick. Some don’t. That’s part of working this out in real time.
The real value isn’t speed — it’s cognitive relief
AI doesn’t make my work faster by cutting corners. It makes the thinking easier by removing background noise.
That’s what I feel most immediately. Not transformation. Not automation. Just less drag on the parts of the job that quietly consume energy and attention.
Next week, I want to sit with the harder part of this conversation: risk, judgment, and guardrails and where communicators shouldn’t use AI, even if it saves time.
If you’re curious, I’m also collecting notes on how I structure boundaries and prompts, not templates, but constraints, for a future edition.
