Today I was having a conversation with three people who are very comfortable with technology and, especially, with AI.
At some point, we started talking about how they organize their meetings, ideas, and tasks.
They were using tools that can record meetings, transcribe conversations, summarize discussions, identify action items, create checklists, and even put tasks directly into their calendars.
On paper, it sounds almost perfect.
But then we started talking about what happens after the meeting.
And that is where things got interesting.
The tools were not always able to understand the difference between something that was actually a task and something that was simply mentioned during the conversation.
A comment could become an action item.
An idea could become a task.
A suggestion could suddenly look like something someone was supposed to deliver.
Sometimes what was needed wasn’t a task at all. It was simply a follow-up conversation with someone.
So, after the meeting, they still had to spend another 10 or 15 minutes going through the notes, filtering what was important, deleting what wasn’t, and figuring out what actually needed to happen.
One person was using one AI model. Another was using a different one. Some had scripts and additional tools to clean things up.
And I looked at them and said:
“You know what’s funny? Didn’t we already have a solution for this?”
The secretary who understood the context
Back in the 80s and 90s, a good executive secretary could do exactly this.
She knew the company.
She knew the priorities of the person she worked for.
She understood the context.
She knew the difference between a comment, an idea, and a commitment.
If someone said, “We should probably look at this,” she could understand whether that meant someone needs to do something or whether it was simply part of the conversation.
She didn’t just take notes.
She understood what needed to happen next.
And that’s the interesting part.
The difference wasn’t better note-taking.
It was context and judgment.
We are now trying to teach machines to recognize something that, for a long time, another human simply understood.
And that made me wonder if we are sometimes approaching these problems backwards.
Maybe the problem isn’t AI
We ask:
What tool should I use?
How can I automate this?
Which AI is better?
How can I make this faster?
But we sometimes forget the most basic question:
What problem am I actually trying to solve?
We start optimizing before defining the outcome.
We choose the tool before understanding the destination.
And sometimes we become incredibly efficient at doing something that didn’t need to be done in the first place.
This isn’t really an AI problem.
It’s a problem with how we approach problems.
And the more I thought about it, the more I started seeing the same pattern somewhere else.
The cost of losing context
Over the last few years, we’ve also changed the way we interact with each other.
More remote work.
More messages.
More online meetings.
More asynchronous communication.
Less sitting in the same room.
Less walking over to someone’s desk.
Again, none of this is necessarily bad.
But there is something that is difficult to replace with software:
context.
When you work closely with someone, you don’t need to explain everything.
You understand their priorities.
You know what they mean when they say something.
You know which ideas are important and which ones are just being discussed.
A lot of that knowledge isn’t documented anywhere.
It is built through interaction.
And when some of that context disappears, we often try to replace it with more technology.
More documentation.
More processes.
More prompts.
More AI.
Sometimes we are using technology to compensate for context that we lost somewhere along the way.
And maybe this is one reason why simple things can start requiring complicated systems.
The illusion of agility
Then there is another habit that makes this even worse.
We constantly move from one task to another.
Email. Meeting. Slack.
Spreadsheet. Message. Back to the spreadsheet.
We call it being agile.
We call it being productive.
But every time we leave a task, we leave part of our thinking behind.
When we come back, we have to reconstruct it.
Where was I?
What had I already figured out?
What was I going to do next?
That transition has a cost.
The more complex the task, the greater the cost.
So what looks like agility can actually be fragmentation.
We feel busy. We feel responsive. We feel productive.
But sometimes we are simply moving faster between unfinished thoughts.
We confuse movement with progress.
Start with the destination
Maybe these are not three different problems.
Maybe they are connected.
We have more tools.
We have less shared context.
We have more interruptions.
And we have more things to manage.
Then we create more tools to manage the complexity created by the tools we already have.
This is why I keep coming back to something very simple:
Before optimizing the process, define the destination.
It’s like a flight plan.
You need to know where you are and where you are going before worrying about the fastest route.
The same applies to our work.
Before asking:
How can I do this faster?
Ask:
Do I need to do this at all?
Before asking:
Which AI should I use?
Ask:
What am I actually trying to accomplish?
And before jumping to the next task:
Is this more important than finishing what I’m doing?
Technology can help us move faster.
But it can’t decide what deserves our attention.
Maybe productivity isn’t about doing more things faster.
Maybe it is about being clear enough to know what doesn’t need to be done, focused enough to finish what matters, and disciplined enough not to confuse movement with progress.
Before we look for another tool, another automation, or another AI model, maybe we should go back to the simplest question:
What exactly are we trying to accomplish
Leave a Reply