Be a Rube.

About 15 years ago when I went back to Ogilvy to help run the IBM account, I came to the realization that while a lot of people jabber on and on about technology, very few people understand what technology actually does or how it actually works.

Even fewer can actually explain it so it’s memorable and sexy.

I also realized that journalism had started doing a better job explaining processes and flows than advertising. Mostly because nobody any more in advertising was dug in on a single piece of business, so no one had the time to try to break things down and show potential customers how they work.

I realize the world was full of blather like the blather below. That might actually be unfair to blather. But the simple truth is that much of what’s explained is instead confuserated and everyone is too cowed to call bullshit on it. This sort of language is everywhere.

I can’t imagine it doesn’t depress sales.

About the time I rejoined Ogilvy, I ran across this article in the Times. 

I had read a book once called “Water for Gotham.” Basically it was about the importance of taking a long-view on infrastructure. It’s the story of how a smallish city with no potable water built a multi-trillion dollar system to bring fresh water from the Catskills to New York’s to quench what in time became 30 million tongues. 

I noticed the Times had told the story with six or seven cartoonish pictures and a few hundred words of copy. The New York water system is one of the most-complex infrastructures in the world. If it can be explained this simply, why can’t we do the same when we’re selling AI, or data migration, or Quantum, or whatever?

As the world has grown more complicated and the need for explanations more prominent, our ability to explain things has grown concomitantly more and more atrophied. 

True, I don’t need to know how an internal combustion engine works or how an electric engine works, but I might want to see some forensics on how these machines can help my business.

In fact the very term “horse power” grew out of such a seed. Almost three centuries ago, people had no understanding of a steam pump. To make it palpable scientists created a calculus—horse power—that explained the amount of work a machine could do in terms people understood. Terms they could also apply as a basis of comparison.

There’s no similar nomenclature of any current technologies. scam altman calls his AI ’super intelligence,’ which is about as meaningless as a holding company’s earning-reports. When new bombs are built we say “it has the power of twenty Hiroshima sized bombs.” That gives people an idea. Why don’t we in tech say something is an MIT⁷, that is, it has the computing power of seven times Dr. Horchner’s advanced physics lab.

We also fail to explain process.

We fail to say, for instance, that a cancer doctor might look in his lifetime at 5000 MRI scans of cancerous cells in a lymph node. An advanced AI system will have been trained on five-million MRI scans. From a volume point of view alone, that can lead to insight and acuity. It allows the AI system to say “that looks off.” It can then summon a human to double-check the system’s work.

If you’re routing a delivery truck through a city grid and there are twelve turns, that’s a relatively simple job. But by the time you have two-hundred trucks in fifty cities making a thousand deliveries each, the possible decisions number in the trillions. That’s what our AI does. Figures out the shortest distance and the fewest number of turns.

On my Kindle, which is about the size of an old TV Guide, I can carry around with me about 2000 books. About one-third the size of Thomas Jefferson’s library, which was the seed of the largest library in the world, the Library of Congress.

One of the books I have on my Kindle is this one. I recommend it to everyone. 

Goldberg’s work is funny.

It’s memorable.

It explains how things work.

We might consider that.

BTW, when I rejoined Ogilvy and IBM, I started keeping a list of articles that took complicated things and explained them simply and memorably. People always kept list of directors and photographers they liked. Why not do the same with data scientists?

Each page of my list has about seven urls on it. The list is 35 pages long. It helps me. 

It can help you.