Sometimes you get lucky. Sometimes you stumble across a thing that leads to another thing that leads to you sitting and watching a two hour documentary you didn't know existed. This used to happen a lot more when the internet sucked way less.
Those clowns at Facebook and Google have a lot to answer for...
A few weeks ago I found myself on the Ghost in the Machine page for the Sundance Film Festival. I was lucky that the festival hadn't started yet and I could buy myself a virtual ticket to watch the film. Lucky because the film was really good. It has been sitting with me for days now like a good documentary should.
This essay isn't about that film. Not exactly. It's more like it's cousin.
In my day job, analytics are pretty important. Understanding the value of something. Was this a good idea or a bad idea? Did this effort succeed and by how much? Should we keep going down this path or try something new? How do these two things compare? Does any of this make sense in light of our overall strategy?
It is a lot of steady analysis. A lot of math. A need for reliable measurements.
Reliable is the important word in all contexts. The least reliable word I can ever count on in any conversation is productivity. The meaningless catch all word used to define human activity in our society.
It stalks us like a predator.
Consultants love to use the word productivity in all their reports. It is how they add weight when talking about value. Each new thing they try to turn into an exploitable engagement gets wrapped up in the word.
Design thinking increases team productivity. The blockchain will increase supply chain management productivity. And of course, right now, generative AI will increase all of human productivity by untold amounts.
The pattern is obvious. So is the grift. You just have to walk into the bargain bin of the book store to see it.
Productivity is a real thing. I can feel it. Almost touch it. I know when it is down. I can hear it when it is humming. And yet, is it definable in a measurable way?
I know I can measure output. The AI CEOs like to talk about this metric. Number of features. Lines of code. Books on the shelf. Cars on the lot. Milk in the store.
Not one of these things mentions value, which is implied in productivity.
The flip side is of course that the tooling will help you think. This will make you more productive. Off load tasks to the machinery to unlock your human potential and all will be well.
Have you ever tried to ask a follow up question to someone who didn't read the book for their book report?
The great grift of the moment, like all the grifts before it isn't about the technology. It's a magic trick. It even has all three parts.
You have your pledge. The normalcy of the work place. The everyday activities people do to get by. Writing code. Designing products. Doing research. Assembling presentations. Engineering machinery. Assembling devices.
Then you have the turn. The magic box that shows these things done so quickly and perfectly. You ask it a question and out comes the answer. Apps are instantly created. Products are on the market. Presentations are made. Medicines are discovered. New machines are designed.
Lastly you have the prestige. The numbers all go up. Your productivity goes up. Your costs go down. And of course, the money just rolls on in.
It is of course an illusion. And like every good magic trick, everyone wants to figure out how the magic box during the turn works. They want to unravel it so they can be the one to say "Aha! I told you the box wasn't magic, just a box!"
I'm more interested in the prestige. The illusion around the measurements is what catches my eye. No matter the magic trick, that always stays the same.
It is a trick within a trick. With words it draws our eyes in one direction, declaring productivity is this glorious activity representing our potential. In reality it measures that activity like we would the actions of washing machine.
It would be impressive if it wasn't so dishonest...
Generative AI isn't a productivity tool. It is an automation tool. A evolutionary leap from RPA, but still task driven. That changes how you measure its value.
And that is where things get really interesting.
What is the value of the investment for a business? At what point does a task completion gain become a productivity deficit? Are our strategic technology partners currently over leveraged in this space?
There is a reason I don't believe in measuring productivity. It is like trying to measure being a person. Only racists believe in that bullshit.
