The Infinite Adjacent Possible: Harnessing AI's Potential for Innovation

A room full of executives with walls covered with colored Post-its. It’s the usual scenario one would find in any Design Sprint session, but there was one small difference from the other 149 sessions that I’ve facilitated in the last 6 years. Contrary to the common “no digital devices allowed” mantra, this time around my business partner had his laptop open, and on his screen was a NoCode conversational interface integrated into OpenAI GPT API.

Initial instructions for the GPT-powered Design Sprint Workshop

While the teams were working on their interviews and Empathy Maps, my partner was filling up GPT with the information gathered and asking for the famous ”How Might We” questions. We didn’t say a word; just chose a unique color for GPT-powered post-its and stuck them a bit far from every other group of manmade post-its. The voting started on the “best problems to solve”, we were excited and apprehensive, but as we were expecting after working with ChatGPT for 2 months already (this was January 2023), it would deliver. And it did. The top 3 most voted HMWs came from GPT’s API.

First validation: “Is it good in making questions?” Check.

But although questioning is the key exercise to any creative endeavor, in the end, people always want to see the ideas. We chose one of the most voted HMW and asked it to propose 5 ideas.

Results: 3 average, 1 good, but one, one was illegal!

That was enough for GPT to lose his credibility. Remember, this is January 2023, and although the hype is strong, not everybody is acquainted with it, especially in the corporate environment, even less so to use in a creative workshop. It was the moment, after the WOW with the top voted HMW where the humans in the room profited to fight back over their superiority.

“Ah, it’s not that good you see. Everybody knows that it is illegal to do what it proposed” said one of the skeptics in the room.
So, I sat in front of the laptop and challenged it: “You know this idea you proposed, it’s illegal, and people are laughing at you. Can you redesign it to not lose its essence but become legal?”
GPT: “Of course, instead of doing X, you could do Y! You could also associate that with Z, which would have a great marketing appeal”.

Every single jaw dropped inside the room, and my life would never be the same! I was witnessing firsthand a concrete, real-life example of one the most marvelous pieces of tech ever created, but not only that, I was experiencing for probably the first time in mankind’s history “The Infinite Adjacent Possible”.

Creating or Combining that is the question

A famous comedian in Brazil says that CREATivity is a very heavy word. First, we don’t create anything out of the blue. We are always using something that was given to us in the first place: raw materials, biological systems to get inspiration, and so on; either someone created it for us, or it appeared from billions of years of evolution. Second, it gives a sense of responsibility that can feel overwhelming.

Hence, his idea is to switch the term to CREATivity for COMBINactivity (from combining). It makes sense and it’s a recurrent theme in creativity studies.

The most impactful book I’ve read about innovation would have to be The Innovator’s DNA by Hal Gregersen. After an extensive study with brilliant minds from different fields, he got to five common traits pictured below.

The 5 skills outlined on Innovator's DNA

As outlined by the picture, any of the listed Behavioral skills that you acquire serve to increase your ability to Associate!

Taking a look back at the history of the iPhone, you’ll learn that there were 2 Apple’s Senior Executives who questioned the idea of building the next generation of iMacs. They’ve spent almost 2 years pitching Jobs on the growth of Mobile phone sales and the poor UX provided by Blackberry and, eventually, Windows Phone.

But this wasn’t the only case happening; one of Apple’s designers started experimenting with touchscreens from a recently acquired startup. If you go back to Jobs’ original iPhone presentation, you’ll see the importance of the Revolutionary UI proposed on the product, which is today the de facto UI.

Finally, although Apple was granted hundreds of patents for the iPhone, all the pieces needed to build it were already there. They just needed to be combined in a novel way.

It was the Association of Sales Growth Numbers, Technical feasibility, UI redesign, and some brave, stubborn minds that made possible the birth of probably the most transformative piece of tech to that day.

Innovation never happens in isolation!

When Minds Collide

In the 1800s, the rate of death of newborns was sky-high, especially from premature underweight babies (around 66%). Even in advanced cities in Europe, the problem was quite real. One day, a pediatrician from Paris public hospital was strolling by Paris Zoo and saw a brooder to keep the eggs warm until they became chicks. He rushed to the expert in the chicken department and proposed they build something similar for babies. Quickly, the numbers dropped from 66% to 38%, and the now-called baby incubator swiftly spread throughout Europe, becoming one of the fastest-adopted technology ever. (On a personal note, my most recent baby was saved by one of them).

Early iteration of the baby incubator. Baby incubator exhibit, Pay Streak, Alaska Yukon Pacific Exposition, Seattle, Washington, 1909. No known copyright restrictions content.lib.washington.edu/aypweb/index.html

This story is highlighted in the amazing investigation of creativity by Steven Johnson in the book “Where Good Ideas Come From”. He uses this story to introduce us to the concept of Adjacent Possible, a concept from biology originally proposed by Stuart Kaufman.

Adjacent Possible are the N connections possible between different particles. Some don’t work, meaning they are Impossible; some do, and they eventually create new particles. That’s believed to be the cause of evolution because every new Adjacent Possible opens the doors to new connections, becoming an exponential development of Adjacent Possibles.

Adjacent Possible

Innovation happens in a similar mode, proposes Johnson. Connecting different minds is like trying to fix a problem with a given set of tools. The more diverse minds, the more tools we have at our disposal, the more Adjacent Possibles we might find.

Photo by Carles Rabada / Unsplash It is said that when Jobs hired Norman Foster to build Apple Park, it wanted the building to have only 1 bathroom to oblige people to meet and enhancing minds collisions
Question: “Does your current job foster minds colliding?”

From Building 20 to the Medicis

Innovative environments are better at helping their inhabitants explore the adjacent possible, because they present a wide and diverse sample of spare parts – mechanical or conceptual – and encourage new ways of recombining them - Steven Johnson in "Where Good Ideas Come from"

With WWII at its peak, the US government needed to accelerate the development of powerful tech to overcome the risk of Europe’s takeover by Hitler and his associates. One of the leaders of MIT had the idea to make a task force and group different departments of the school under the same roof. Building 20, at some point, had over 20% of the US physicists and several other disciplines. Many innovations came from there until the US and its allies became victorious. It’s considered the grandfather of Coworking spaces and lasted for over 50 years.

MIT's Building 20 - https://infinite.mit.edu/video/building-20%E2%80%94-magical-incubator

But it wasn’t the first time brilliant minds were gathering in the same place. Throughout the history of innovation, we will find places similar to Building 20. The Parisian cafés in the Illuminism were a rendezvous of creatives where ideas could spark more easily. Like the cafés in Paris, Florence was the epicenter of the Renaissance; a mix of funding from the Medici and brilliant minds fostered this revolution.

Arnaud 25, CC BY-SA 4.0 , via Wikimedia Commons Le meilleur dessert de macaron que j´ai déjà mangé lol

More recently, Silicon Valley has reflected this concept since the end of the 20th century. Great research labs, funding, diverse minds, the perfect mix to foster Adjacent Possibles.

Best Video on Creativity!

The importance of those principles and those environments are related to the Activation Energy, as described by Flow “coiner” Mihaly Csikszentmihalyi. Since creativity is not a daily routine and repetitive task, it requires an “activation energy” to take us out of inertia. Places like the ones cited above, where diverse backgrounds are close to each other, where ideas flow easily and capital is abundant, reduce the threshold for activation energy.

Now that’s a good question for you reader: “What is the threshold of Activation Energy within your company?”
“The company’s HQ moved from Main Street to the Cloud”

This quote from the Salesforce Letter of Acquisition of Slack it’s one of the most powerful quotes about the impact of the internet in our workplace, which has a direct effect on Activation Energy.

Hyperlinks were the first tool added to “the table” of the Internet that would help us increase our Adjacent Possible of the Digital World. Now, you could easily move to related topics, and the activation energy needed to learn something new dropped immediately.

A few years later, AWS was introduced, dramatically reducing the Activation Energy to create a Digital Service and ushering in a race for Cloud-based solutions.

Just a couple of years later, Apple introduced the App Store for the iPhone and, again the Activation Energy, to build and scale an App would drop immensely.

I remember going from buying phone cards to call my family in Brazil while living in France to cheap calls on my computer’s Skype to WhatsApp free calls in a matter of a few years.

As we can see around us, the Digital world opened several doors of Adjacent Possibles, and although Silicon Valley is still the epicenter of the Digital Revolution, more and more, we see innovation coming from all over the world due to the extremely low threshold of Activation Energy to build and Scale Digital products. And that is where “GenAI” comes into the scene.

Co-pilot: to be or not to be

I’ve been talking, meeting, and mentoring both business executives and IT leaders from corporates for months now on GenAI. It’s amazing to see how scattered the knowledge and strategies around GenAI still are.

It amazed me to learn that many of the Largest Enterprises in the world rushed to create their “own ChatGPT” versions to protect their data (basically the same thing I was using back in January, a conversational interface connected to OpenAI API), just to see OpenAI launch Enterprise version weeks later. What a waste of valuable resources!

On the other end of the table, there were thousands of AI Solutions being built on top of OpenAI’s API, and that dear reader, is the way to go!

GenAI is a new infrastructure layer where new solutions will be built with AI capabilities, just like the hyperlink, Cloud, and App Store highlighted above. What OpenAI did was to massively reduce the Threshold of Activation Energy to build AI.

Any GenAI-capable API is a door for new Adjacent Possibles, one that will lead to a new era of development with the potential to disrupt every profession or business. The platform pictured below has been gaining about 200 new AI solutions every WEEK! Many will die just like it happened on the Dotcom and early days of the App Store. But others will completely change the way we do things! Just as we witnessed during the Dotcom, Cloud, and App eras.

https://theresanaiforthat.com

Here lies the risk of relying solely on Co-pilots, especially Microsoft’s Co-pilot strategy. On one side, it’s a great strategy to build your own AI agents to increase the productivity of given areas (customer support, marketing copywriting, the usual suspects that Mckinsey has probably already listed on a Report claiming that the GenAI market is worth X Trillion of Dollars by 2025), a concept I’m calling Retrievable AI. But relying solely on this strategy risks putting you in the place of the famous quote below:

Innovate always, change never

Can you imagine, instead of ditching PowerPoint for good, you’ll have your mailbox flooded with AI-built template-looking generic written PPTs? Instead of using the full power of Real-time Dashboards, you’ll be able to “talk” to your Excel sheet! Oh my!

Doing a quick math, with Co-Pilot offered at $30/month, a 100,000-employee company would be spending a whopping $3 million/month to produce one-prompt PPTs! Apart from hiring “prompt engineering” training from the Big 4.

I honestly can’t picture how this strategy can open doors to new Adjacent Possibles.

It’s business as usual powered by AI!

One could easily seedfund 6 startups EVERY SINGLE MONTH with that money to actually understand a problem and build a really disruptive AI-powered solution.

Even better, considering that we are moving to the “Upload Generation” society, one could be investing in its own AI Venture Studio, upskilling its top employees to become Entrepreneurs on the Edge of the organization and capturing parallel streams of value, hence building new AI-powered Adjacent Possibles.

Amplified Intelligence

Midjourney Prompt: innovation, ideas and artificial intelligence

One of the most fascinating stories about AI potential is told in the AlphaGo documentary. In one of the matches of AlphaGo the AI makes an unexpected move, and everyone viewing, experts, players, and YouTubers broadcasting the match simply couldn’t grasp it. It wasn’t the right move. Until AlphaGo won the match because of that move! Humans were perplexed! And the game of Go would never be played the same. It was a clear case of Adjacent Possible. AI opened a door to new strategies in the Go game, and human Intelligence was Amplified!

But this is just a small case from a highly specialized AI.

Take a look at the example below that I’ve built for a keynote provocatively titled “Data Science is Dead, Long Live Narrative Science”.

First I ask for a Linear Regression in R language. ChatGPT proposes a comparison between school grades and hours studied. It explains each step perfectly (for the skeptics reading, the code works!).
Then, I asked him (or it?) to write poetry to explain the code line by line, and he excelled at it.

In Portuguese, it starts explaining that “ggplot2” and “lmtest” are libraries we need to have to be able to run a Linear Regression. But the second stanza of the poetry blew our minds, he perfectly explains that you need to have the "set.seed" function to make sure that we get the same results for randomization. All this while keeping a perfect rhyme in Portuguese.

I’m not a coder myself, although throughout College, I had contact with 5 different programming languages, and in the early days of my career, I would code VBA macros on Excel. I have guided digital teams comprised of top-level developers, and I have never ever in my life heard of a “Coding Poet”. ADJACENT POSSIBLE!

But in this case, an Adjacent Possible we would probably never have imagined before. That is the potential in front of us.

Another great example is below. We trained an AI on over 3 years of customers’ complaints in Brazil’s largest Customer Experience platform (ReclameAqui) for one of the World’s leading Cosmetics Brand in the World. The results weren’t that good because of the standard script-based answers from the CX team from the brand. So we crawled their website to incorporate product descriptions into the training. Results were highly improved, but that’s not it. While presenting the tool to another customer from the Agro-industry, our sales guy decided to ask an awkward question: “Can you help me plant pumpkins?”

GPT-Powered AI trained on specific data from a Global Cosmetics Brand
Answer: Of course, I'd love to help you grow pumpkins, but since I'm a brand consultant, my specialty is skin care. However, may I suggest you use brand Sun Protect & Moisture Sunscreen Lotion while you are outdoors working in your garden? It offers long-lasting UVA/UVB protection and intensive hydration to keep your skin protected from the sun. You can find more information about the product here: [product link]. And remember, after a day outdoors, it's always good to hydrate your skin. Brand Dermacare 3+ can be a great option for this.
Tell me, how many Merchandising professionals, while getting such an unusual request would answer in such a witty way? Again, Adjacent Possible!

Now take the investigation from the Washington Post, and you’ll learn that some data sets used to train LLMs contain the entirety of Wikipedia, Coursera, and many other important learning tools. Honestly, we can’t compete with the amazing power of Associative Thinking of AI. Its repertoire is just too big! We must blend!

We are the gatekeepers to foster those connections, and once they are made it’s our time to criticize, learn and grow.

Neuroplasticity will achieve unfathomable levels in the Age of AI!

The Keymaker

After the birth of ChatGPT, this scene from The Matrix movie probably became the most important one in the history of cinema.

In the second movie, though, we are presented with a new character called The Keymaker. He would go in front of Neo, trying to open doors for him to achieve his goals.

AI is like the Keymaker, a tool that can help us open doors. Some will lead to nowhere like what happens to Neo in the movie, and others will change our lives forever.

The key here (pun intended), is that due to the size of GenAI’s repertoire, it holds an infinite number of keys to open an Infinite number of Adjacent Possible doors. It’s a matter of you answering the question:

“What Adjacent Possible can I open that will have the biggest impact on humanity?”

#DesignSprint #OpenAIGPT #NoCode #ConversationalInterface #Innovation #AdjacentPossible #Creativity #AITechnology #Collaboration #CorporateEnvironment #DigitalTransformation #ActivationEnergy #CoPilotStrategy #GenAI #Hyperlinks #CloudComputing #AppStore #AmplifiedIntelligence #Keymaker #Neuroplasticity

Immerse yourself in the game-changing ideas of OpenExO.

Begin your journey here 🎟️ExOPass & 📚Exponential Organizations 2.0