View profile

Thinking Askew – The "How I Got An AI To Write A Dadaistic Article About Innovation"-issue

Thinking Askew – The "How I Got An AI To Write A Dadaistic Article About Innovation"-issue
By Alf Rehn • Issue #21 • View online
Hello, friends! So, one of the most talked about machine learning-projects at the moment is GPT-3 – “an autoregressive language model that uses deep learning to produce human-like text”. It is a very powerful engine, and it has been used to create texts that sound an awful lot like a human being created them. At least when it is well deployed, that is. Today’s Thinking Tuesday is a story about what happens when the cracks start to show…

I Wrote An Innovation Think-Piece With Machine Learning, LOL!
Sooo… I decided to take GPT-3 out for a spin. More specifically, I signed up for a service that promised to write content with this magical-sounding not-quite-AI-but-not-quite-not-AI-either thing. My thinking was that I often find that innovation think pieces sound like they’ve been written either by a robot or by cutting and pasting, so I thought “Great! I can have fun with this!”. In the back of my mind, I already thought that maybe, just maybe, I could have it do a bunch of writing for me. So I fed the service (which did point out that what I was doing was going to generate weaker results than in the niches they’d pre-fed with materials) with some text of my own, gave the engine a nudge, and waited for what I expected to be a vapid text, but one scarily like many content marketing posts about innovation. How wrong I was. Below, in its grandiose entirety, the essay that was generated, with odd capitalizations and other idiosyncrasies kept intact. At the end, I’ll write my commentary. Hang on, it’s going to be quite a ride!
Sustained myths about creativity and innovation lead to confusion, bad practice and bad decision making.
So, let’s take a look at five factors that disguise Creativity and Innovation as hope and one that most often do more harm than good.
1. The Art of the Impression
“Getting to yes is a new psychological or cognitive technique. It is the art of the impression or impressionist technique.”
~Thomas Sowell, The Fractured Republic,
I find it hard to take any person or organization very seriously when they reference a point of view about what it means to be creative; perception is the first step to true Innovation. We can be creative all we want, the best tool we have is the eye and imagination. As project managers, managers and business writers, we often get caught up in the fast moving idea flood of an unpredictable world. Companies get away with behaviors that actually undermine the success of a company or group. What is the outcome of this?
Imagine a series of blunders are made in an attempt to find a way to solve an issue, create a viable idea and get the approval of management. This leads to insufficient time to actually develop an idea or concept. As a consequence, money, time and reputation are lost.
The key to innovation is: find the right fit to activate an existing existing or future thing. An idea goes into the black hole of the information wasteland because it does not match the mold of an existing business. Project managers spend years running scenarios and conducting research, searching for an idea they can fund with the money they have allocated. The same happens with managers. Machine historical research plays an important part in the creation of an idea in many industries, including marketing, finance, engineering, design, activities and more.
In recent history, it seems we have been brainwashed to see shows like NCIS and CSI as an unmissable entertainment 24-7 and love what we do despite the negative impact it has on real people. Like all the other likes of “brain training”, the odd man out business model has finally cracked into the creative industry.
Let’s take a look at a few long-term examples of What not to do.
1. Think the Next Question is The Next Question
“Break bad habits. This can be done using a combination of grinder or mincing shoe stumps (the punch bowl in the toolbox is great for processing the first few steps); and learning to ask the right questions: Does this need a funny nickelcoball? Is this keychain engaged in crime? Is this a fun story for my visitors to carry at key moments?”
~Carlos Larson, JUNAPULATE!,
Many companies project agendas onto the outside. The core is the importance of being real; asking what motivates people. Answering these questions will immediately show that the project is not actually an overall improvement (Things that go well can be built onto the side). If the idea includes a side bet, it is no better than the bet in the first place.
Let’s say I want you to build the next #happens on the road. Well, that is perfect, but were designing the next theme park or this new way for our business to talk to our customers and make it simple. No. We need to solve a radical problem. The solution must take current business logic to transform it into an industrial design product.
Scenario: We raise money from investors, the idea is successful, the next step involves the right team from the right district for this country. It will be a long tough journey, but if nothing else the idea deserves our oxygen. But the whole journey will not stop there. The great thing about a concept that truly takes off is that we will be able to create some real capital to help us create an actual product.
Realize that we have already put in an investment, and we not only need to use our income to buy the right product that solves a horrible problem but also to kick start another revenue towards the next set of growth over the next several years.
Last year, I had just entered active duty, next year I expect to be way back on the road. We are doing “What happens on the road” a mile a minute. If this next commercial idea finally works, it will increase my income for kickingstart our growing past operations, so our cash flow is protected through the post deployment period. So, remember the Friend, Family or the current employee for all those investments you are making.
Instead…
2. Do Not Pile Up More Data
“Data and calculations are just the first step to the solution. It is the two-step process of integrating data and decision making. How much data do you have? How much can you use? How much data does it require to successfully embrace change? How much cost?”
Alf’s comments
I don’t know where to start. Five factors, count ‘em: 1, 1, 2. There seemingly is an industry of activities. What about that really odd shoutout to NCIS and CSI? Nickelcoball! Ideas seemingly deserve our oxygen, and both kick starting and kickingstart are important. Oh, and what about those companies with behaviors that undermine the success of companies – is the algorithm trying to describe competition, or something more existential about capitalism?
Then again, there’s some really gems in there. I don’t know about you, but I really want to read “JUNAPULATE!” now, if only to know whether that keychain engaged in crime (I really hope it did.). Also, #happens sounds like the next big thing on TikTok. “Don’t pile up more data” is often good advice, as well. Oh, and I sort of want to know what the odd man out business model is…
My takeaway? I think innovation writers might be safe for a little longer. Still, this little experiment raises questions. Why did GPT-3 make such a hash out of the task of writing an article about innovation? Is it because questions of creativity and innovation are something that only humans can fully grasp, or is there something more ominous at play? Might this be a case of garbage in, garbage out? Might the mighty machine learning engine simply have been stumped by the word salad and vapid declarations that plague innovation writing, and simply not found enough sense in there to emulate? Might the text above be the sign of innovation writing to come? Time will tell…
That’s it for this time! Share if you enjoyed it, or drop me a comment. Stay safe out there.
Did you enjoy this issue?
Alf Rehn

Professor of management & innovation, speechifier, and popular culture geek.

In order to unsubscribe, click here.
If you were forwarded this newsletter and you like it, you can subscribe here.
Powered by Revue