A few months into using AI properly, I started keeping a rough log.
Not formally — just a note on my phone when I noticed something. A task I’d finished in ten minutes that used to take an hour. A thing I’d been putting off for days that turned out to take twenty minutes once I stopped trying to do it entirely on my own.
After a month I had a list. Here are five things from it.
1. Summarising a Long Email Thread
Before: twenty minutes of scrolling, rereading, trying to remember what had been agreed and by whom, occasionally going in circles.
After: I open the thread in Outlook, click the Copilot icon, and type:
“Summarise this thread. Tell me what’s been decided, what’s still open, and what I need to action.”
Two minutes. Sometimes less. The summary isn’t always perfect — occasionally it misses something I have to go back and check — but it gives me a map of the thread in seconds, which means I spend my time on the parts that actually need my attention rather than the archaeology.
2. Turning Meeting Notes into a Summary
Before: forty-five minutes minimum, often done badly because I was tired and it felt like admin after the fact.
After: I paste my raw notes into Copilot — messy, abbreviated, half-sentences — and type:
“Turn these into a clean meeting summary with agreed actions, owners, and deadlines clearly listed.”
What comes back needs editing. The owners aren’t always right and I sometimes have to fill in context it couldn’t have known. But the structure is there. The thing that used to feel like a chore now takes fifteen minutes including the edit.
3. Drafting a First Response to a Tricky Message
Before: I’d read the message, feel the mild dread, close the email, and come back to it later. Repeat. Sometimes for days.
After: I paste the message into Copilot with a brief explanation and type:
“Draft a professional response that acknowledges their concern, addresses the main point directly, and keeps the tone constructive.”
I’ve written about this before — it’s the blank page problem. AI gives you something to react to. Even if I rewrite most of what it produces, starting from a draft is categorically easier than starting from nothing.
4. Researching Before a Meeting
Before: a slightly anxious hour of googling, opening fifteen tabs, and trying to synthesise something useful from a dozen different sources I hadn’t chosen very carefully.
After: I open Perplexity — which searches the web and gives you sourced answers rather than just text — and type:
“Give me a briefing on [company / person / topic] ahead of a meeting. Include recent news, key context, and anything I should know before walking in.”
Five minutes. I verify anything that matters independently, but the starting point is solid and I walk in prepared rather than slightly panicked.
5. Preparing Talking Points for a Difficult Conversation
Before: I’d run through scenarios in my head, usually at eleven o’clock at night when I should have been asleep, never quite feeling ready.
After: I describe the situation to Copilot — who I’m meeting, what the issue is, what I want the outcome to be, what I’m worried they’ll say — and type:
“Help me prepare for this conversation. What are the key points I should make? What objections might I face and how should I handle them?”
What comes back is a starting point, not a script. But thinking through the likely objections in advance, in writing, is genuinely useful — and doing it with AI means I do it properly rather than running it on a loop in my head.
The Thing I Actually Noticed
Looking at that list, the tasks I was doing in an hour weren’t difficult. They were draining. The hour wasn’t full of hard thinking — it was full of friction, false starts, and the low-grade resistance that comes before doing something you don’t particularly want to do.
That’s what AI addressed. Not the expertise. The energy.
The hour I saved isn’t an extra hour of work. It’s an extra hour of arriving at the things that actually matter with something left in the tank.
That, more than anything, is why I kept using it.
— Anna