Episode transcript The transcript is generated automatically by Podscribe, Sonix, Otter and other electronic transcription services.
Hello everyone, here is Ricardo Vargas, and this is the 5 Minutes Podcast.
This week, I recorded a video for a course on LinkedIn, and one story has stayed in my head since then.
It is the story of electricity arriving in factories at the end of the 19th century.
At that time, factories were built around a big steam engine.
The engine turned a huge shaft that ran along the ceiling, and that shaft moved every machine through belts.
That is why factories had several floors, all stacked around the steam engine.
The place of each machine depended on its distance from the shaft, not on the logic of the work.
When the electric motor appeared, most factory owners did what seemed obvious.
They removed the steam engine, put an electric motor in its place, and kept everything else.
The same shaft, the same belts, the same building. And productivity barely changed.
In 1899, electric motors provided less than 5% of the mechanical power in American factories.
The big jump in productivity only came in the 1920s, about four decades after electricity became commercially available.
The economist Paul David, from Stanford, explained this delay in a paper published in 1990.
The gains only came when someone had the courage to take the shaft off the ceiling.
Each machine got its own motor.
The factory moved to a single floor, organized around the flow of the product.
And the change was not only in the building, because people started making more decisions and had to learn a new way of working.
I am telling this story because I see the same thing happening today with artificial intelligence in our projects.
And I am not talking from a distance. I do this too.
Think about the status report that used to take a whole morning to prepare.
Today, with AI, it is ready in a few minutes.
The meeting minutes write themselves.
The schedule summary arrives ready in the sponsor's inbox.
This is progress, and I want to be very clear that I am not criticizing it.
Saving time is good, and it deserves to be celebrated.
But if we look closely, what we did was change the motor.
The shaft is still up there on the ceiling.
So I want to ask a question that bothers me a lot.
Have we ever asked ourselves whether the report will exist in the future?
The status report was born for a very simple reason.
Information did not reach the people who had to decide on their own.
Someone had to collect the data, organize it, summarize it, and carry it up.
The report was the shaft of the project.
Now imagine a sponsor who can ask for the project data directly, at any time, what is late and why.
In that world, making the report ten times faster is like putting a stronger motor on a shaft that may not need to exist anymore.
The same question applies to many things we do every day.
The weekly meeting that only passes information along.
The approval that goes through four levels just to confirm what everybody already knows.
The committee that meets to read out loud a document that everyone received the day before.
None of this is wrong. All of it was designed for a world where information was expensive and slow.
What does this change in our work as project managers and sponsors?
First, when we build the business case for an AI project, we need to separate two kinds of benefits.
One is time saved, which is changing the motor.
The other is a redesigned structure, which is taking the shaft off the ceiling.
The first one shows up quickly and is easy to measure.
The second one takes longer, demands more work, and touches power, roles, and habits.
But it is the one that really makes the difference.
Second, we need to be careful with the time that AI frees up.
If nobody decides where that time goes, it is usually filled with more of the same: more reports, more versions, more meetings.
Then we change the motor and hang even more belts on the shaft.
Third, there is a simple question that helps a lot at this point.
If we were starting this process today, from scratch, with the technology we have now, would it exist the way it is?
If the answer is no, automating it may just be a faster way to stay in the past.
So here is my challenge for this week.
Pick a single artifact from your project, for example, the weekly report.
Ask which decision it supports, who makes that decision, and whether that person still needs the report to decide.
Maybe the answer is yes, and that is fine.
But maybe you will find that you are keeping a shaft on the ceiling just because it has always been there.
The factory owners who only changed the motor were not naive, because replacing a building that still worked was expensive.
But the big gains went to those who had the courage to redesign the building.
Maybe the most valuable AI project in our organization is not the one that produces the report in ten minutes.
Maybe it is the one that makes the report unnecessary.
I hope you enjoyed this episode, and see you next week with another 5 Minutes Podcast.
