Ford vs Ferarri & Information vs A.I.
Ford vs Ferarri, Sabermetrics vs the Seeing Eye. Should we spend our time teaching a computer how to say “hello?”, or spend our time taking 2,000 years of global, human information and start making sense of it?
My colleague, Steven Birdwell and I were having this discussion after we saw the Network of Words, aka(NOW) that our company had developed. It had just executed a series of really interesting use cases. We were chatting on our way to get some food and were getting a bit abstract and philosophical.
This is because the fundamental idea behind this is; taking existing human information, cataloging it, and applying that to real-world issues and solutions, like Supply Chain, S&OE, Product Transparency, Product Planning, amongst other things. What it does, if used correctly is allow one to see and connect dots unseen before.
For example, by gathering data, and turning into information, via publications, articles, videos, images, the Network Of Words is able to pinpoint Ripple effects. Most recently, for example, the Ripple effect that the coronavirus has had on a global supply chain or even a networking event.
So what of A.I.? Well, I’ve always felt computers to be stupid. Having studied computer science, I know how difficult it is to tell a computer a series of instructions so that a given application works. Behind the inner workings of the computer, big or small was coded up and written by a human being.
To me, there is no replacement for a human being. If I’m being blunt, coding is difficult because computers are tough learners. If it were sitting in a classroom, it wouldn’t be the brightest bulb.
My brain, your brain, and certainly our collective brains are more powerful than a computer. Why not leverage the millions of books, articles, and human entered information, whether it’s structured data, like a book, or unstructured like a tweet, to fill the cracks that exist in every enterprise?
If you want to create a real-life version of iRobot, then I think A.I. might be your cup of tea. But for the rest of us, that are reasonable, and simply want to provide actual value to our fellow businessmen and women, then we all must think differently.
In order to change the way we look at the impact of information, we have to to look at how it is all managed and stored. Right now, information is turned into ‘data’, that is then shoved in some sort of SQL database by some entry-level engineer or intern. Then some other intern has to come along, write a script using Python and retrieve that data, make sense of it, and convert it back into real information that makes sense.
First of all, what the f? Am I the only one that sees the redundancy?
I bring that use case because I used to be that intern. I used to enter and retrieve data into and from a SQL database and turn it into back information, and vice versa at Barracuda Networks, a network security company. Now, I learned a lot. My primary computer science skill is probably understanding databases. SQL and these ‘databases’ are simply an intermediary asset that allows for interactions between a human and data storage.
I wonder if Michelangelo and the Sistine Chapel come together because of a data?
If we used information instead of data, could we protect the integrity of creation that touches nearly every industry.
I spent the last year, 2019, breaking down the product lifecycle of a given garment in fashion. It starts with common sense. You brainstorm and look for inspiration. You then design a rough initial design. So far, not too different than the development of a digital application. You then get a sample of the design, aka a beta or test version of a given application. There is an internal approval process that commences. Again, not too different from what we typically do in software. The only difference is that the information is never converted into data. It’s cataloged. The information regarding the inspiration is cataloged. The construction, the fabric, the style of the lapel or a cuff or the buttons live in a simple cataloged excel spreadsheet called a ‘Bill of Materials’ or BOM.
Now inspiration can come from anywhere. Unstructured data comes at you while scrolling through your Instagram feed where there is something that catches your eye, or blogs published to Medium like this article right here, to 'high end' unstructured data that comes from aesthetically and content-rich books like ‘The Italian Gentleman’ by Hugo Jacomet, ‘Tom Ford’ by Tom Ford and Bridget Foley, ‘The Best Menswear in the World’ by Simon Crompton, or ‘Elegance’ by Bruce Boyer.
What if we were to catalog the information in unstructured information sources like those books and combine it with structured data, ask a question and get answer?
I actually worked on something similar, except I was using Twitter to figure out demographics and marketing data. If the project hadn’t tanked for various reasons, the tool, to my estimation, would have filled in a lot of grey area when it came to marketing, predictive sales information and more.
All of that information would have been compiled from simple tweets by human beings. My conclusion after that project is that it is up to our intelligence to make the best dissertations using what already exists. The power of human intelligence. Not artificial intelligence. Let’s look at the technology that can supplement our lives, as opposed to a replacement.
I just watched Ford vs. Ferrari. I’m going to use this as an analogy. Can you guess which car was built using simple human intel and intuition, and which one was designed using a data-driven system?
The Ferrari reps escort the Ford gentleman through the factory, where he explains that each component is hand-built by one man, each transmission is hand-built by one man and so forth.
After numerous humiliations, Henry Ford II came to Carrol Shelby, who at the time ran Shelby American, and was the only American to ever win Le Mans, with one demand: Win, under the Ford guise.
They threw out the data-driven playbook that Ferrari, Porsche, and even Ford used for their regular line of vehicles. This was that period's version of A.I.
The car that ends up winning is Ford. Ford, the way the movie portrays it, essentially throws away the book, going for common knowledge instead, that they, as experts and gear heads knew, like the back of their hand.
There is a scene where they are just chucking out “unnecessary” luxury features, like the audio system, unnecessary upholstery, and just using the spirit and knowledge of racing as inspiration to make a car faster.
What if that knowledge was available?
Well, it is. A good amount of it is in Carroll Shelby’s own authorized biography. And more are in books like “Go like Hell: Ford, Ferrari, and Their Battle for Speed and Glory at Le Mans”.
Let’s not forget about our hero, and driver, Ken Miles, who put his life on the line and seemed to understand what made a car tick better than anything in the world. This is knowledge. This is information. No data.
A.I. is trying to put that common sense into a computer. It’s just not possible.
What is possible is to fuse the two. To use human information, with the assistance of data.
Similar to what happens in Major League Baseball.
It used to be that you had a group of scouts, a manager, and so forth, that used goggles and a big mouthful of chew to analyze a player's worth. There is no denying that there is no replacement for the human gut instinct. Like Brian Sabean, a young scout for the Yankees who demanded that New York draft Derek Jeter. Or later on in his career, Sabean used his gut instinct and naked eye to draft Buster Posey who would lead the club to 3 world champions, a rookie of the year, a league MVP, a batting title, and so much more.
Sabean was old school. He didn’t believe in analytics in Baseball. Meanwhile, across the bay, Billy Bean, the man operating the Athletics took a completely different approach. An approach that would lead to a book and then a movie called Moneyball. While it was an interesting business story and a good movie for Hollywood, the Athletics have not won squat. They have seasons where the analytics come together but they eventually fall apart, because a computer can’t tell you to sign Josh Donaldson. A player they traded in the early 2010’s, that eventually led to their downfall.
Still, other clubs were taking a hard look at what the A’s were doing. There had to something this method. The A's had the lowest payroll in baseball but were competitive. About every 4 years to the tee, they would make a solid playoff run.
The A’s had, and still have the lowest payroll in big league Baseball yet still made the playoffs more than they statically should have.
Alas, teams started to combine what we now call saber-metrics (deep analytics) and the old school seeing eye to make scouting decisions. When adopted by teams with money like the Red Sox, Yankees, Giants, Dodgers, Cubs, and Pirates, you suddenly start to see titles. Even Brian Sabean, who reminds me of a mafia boss, was open to at least trying it out.
The Athletics meanwhile seem to continue to operate as an artificial intelligence agency.
In using the Network of Words, a human is putting their common sense into a computer, visualizing it as a mind map, and allowing themselves to research and correlate information in a way that was never possible before. To see connections between two independent pieces of information, and gaining insight from these connections. It is the tool that will allow you to see the dots unseen before…