How AI Kills Linear Thinking

Linear & Non-Linear Thinking

Let’s start with some definitions:

“Linear thinking is a systematic and analytical thought process that follows a known step-by-step progression similar to a straight line. Linear thinkers view a problem as a process with a set starting point that follows a sequence of connected series, ultimately leading to a solution … thoughts flow in a straightforward, logical way, and progress in a stepwise fashion. Linear thinkers are considered left-brained because the left side of the brain is associated with methodical and analytical thinking.”

“Non-linear thinking is less constrictive – letting the creative side of you run rampant because of its inherent lack of structure. It’s kind of like letting a puppy run wild on a walk up a mountain – anything of interest will be thoroughly investigated (and perhaps peed on) before jumping to the next, possibly non-related subject! It’s very much like brainstorming – allowing thought to flow, unhindered, in attempts to arrive upon something special in the process … non-linear thought increases possible outcomes by not being so certain about the starting point for any logic process.”

Note that there are variations on these definitions, including lateral thinking, systems thinking, design thinking and fragmented thinking – among others – but for our (and Gemini’s) purposes here the simple linear versus non-linear distinction is fine.

Always Contextual

Linear and non-linear thinking can be good or bad, depending on the problem-solving context. Linear thinking is generally “easier” because the starting point is known. It’s also the stuff of incremental thinking and from a business POV safer, easier to understand (and therefore more defendable) and far more predictable (though of course it can be wrong). Most managers and leaders are linear thinkers: we’ve all heard insults hurled at the boss for being a simple linear thinker!

Listen to this:

“Linear thinkers’ brains are wired differently than non-linear thinkers. As engineers (software, chemical, mechanical or electrical), information technology specialists, accountants and other research-based professionals, their linear style of thinking generally serves them well in their chosen professions. However, that can change as they move up the chain of command to leadership roles.” It can also change when a new technology arrives on the scene with its own kind of “thinking.”

Thinking & AI

Lots of well-bounded, linear processes can be performed better, faster and cheaper with simple machine learning tools. Regression + supervised learning is AI’s little secret: they’re everywhere and they work. But the game-changing opportunities lie with generative AI and its potential for non-linear thinking and problem-solving. The leap from supervised to unsupervised learning (thinking) is one way to think about the potential. But it goes way beyond simple applications. According to Gemini — in response to the prompt – “are the applications of generative AI more linear or non-linear?” – here’s how to think about the potential:

“The applications of generative AI are driven by non-linear models.

“Here’s why:

“Real-world data has complex relationships: Generative AI deals with things like images, text, and music. The connections between features in this data (e.g., colors in an image, words in a sentence) are intricate and often context-dependent. Linear models struggle to capture these complexities.

“Neural networks are the workhorse: Generative AI relies on neural networks, which are inspired by the human brain. These networks use non-linear activation functions to process information and learn patterns. This allows them to handle the nuances of real-world data.

“Examples highlight non-linearity: Consider generating a new song. A linear model might just string random notes together. A generative AI, however, can learn the complex relationships between melody, harmony, and rhythm to create something that sounds musical.”

I then asked Gemini: “what are some linear business problems that AI solves very well?”

“Linear business problems are a perfect fit for AI’s strength in optimization. Here are some examples of how AI tackles linear problems in the business world:

• Inventory Management
• Pricing Optimization
• Route Planning
• Marketing Campaign Optimization
• Financial Ratio Analysis

But what about ROI? What if a company wanted to know which flavor of AI it should invest in? I asked: “how do I get the best return on my GenAI investments? Linear or non-linear business problems?”

“GenAI’s power lies in tackling non-linear business problems. Here’s why focusing on non-linear problems offers a higher potential return:

“Unlocking Hidden Value: Many complex business challenges involve intricate relationships that linear models can’t capture. GenAI, with its non-linear approach, can identify subtle patterns and connections in data, leading to innovative solutions and uncovering hidden value.

“Boosting Creativity and Innovation: GenAI can generate new ideas, automate creative tasks, and personalize experiences. This can lead to entirely new product concepts, marketing strategies, or customer service approaches that wouldn’t be possible with linear thinking.

“Augmenting Human Expertise: GenAI can analyze vast amounts of data and identify patterns that humans might miss. This can empower human employees to make better decisions, solve problems more creatively, and focus on higher-level strategic tasks.

“However, this doesn’t mean you should ignore linear problems entirely. Here’s a combined approach for maximizing your return:

“Identify Bottlenecks: Start by using AI for tasks well-suited to linear models, like optimizing pricing strategies or managing inventory levels. This frees up resources and provides a solid foundation for further investment.

“Focus on Non-Linear Opportunities: Then, explore how GenAI can address complex business challenges with non-linear elements. This could involve personalizing customer experiences, developing new content formats, or automating design tasks.

“Prioritize Strategic Impact: Align your GenAI investments with your overall business goals. Focus on areas where AI can create a significant competitive advantage or unlock new revenue streams.

“By combining the efficiency of linear AI with the innovative power of GenAI, you can maximize your return on investment and drive real business transformation.”

Obstacles

The investment strategy is clear, but since most managers and leaders are linear thinkers (and have a limited understanding of generative AI), how big is the problem?

“Business professionals are often trained and rewarded for linear thinking. This can make adopting non-linear thinking for GenAI applications challenging. Here are some specific hurdles they might face:

“Comfort Zone: Many professionals are comfortable with the predictability and structure of linear models. Shifting to a more fluid, exploratory approach of non-linear thinking can feel uncertain and risky.

“Metrics and Measurement: Measuring the success of non-linear solutions can be tricky. Traditional metrics focused on efficiency and cost reduction might not capture the full value of creative solutions generated by GenAI.

“Lack of Experience: Working with GenAI requires new skills and knowledge. Businesses might lack the internal expertise to fully utilize the capabilities of these non-linear models.

“Bias Towards Established Solutions: Existing processes and solutions, even if suboptimal, can be ingrained in company culture. Business professionals might be hesitant to embrace new, non-linear approaches from AI.

“Fear of Failure: Non-linear thinking can lead to unexpected outcomes, some of which might not be successful. Risk-averse professionals might be hesitant to embrace this possibility.

Next Steps

First, prioritize non-linear thinking and investments. Entrepreneurs should especially think this way. Pascal Finette says it plainly:

“Linear thinking is the path to doom for new ventures. We’re losing lots of opportunities in a world where linear thinking clashes with exponential trends.”

Machine learning and generative AI offer enormous opportunities to save money and make money. In order to identify and optimize these opportunities, companies need to search and destroy embedded linear thinking. While this does not mean there’s no payoff to linear problem-solving – which there clearly is – it means that in order fully leverage generative AI new approaches to “thinking” must be adopted. Inevitably this means jettisoning the old ways, which is perhaps why Marriott did not “invent” Airbnb or VRBO. Or why McDonald’s did not invent Impossible Meat or how to increase its appeal or lower its production costs instead of abandoning the whole McPlant initiative. Or how Uber came to be? Or why autonomous agri-business is yielding great returns?

Is non-linear ultimately about innovation? Absolutely, but it’s also about the marriage between innovation and generative AI. Ask Gemini about “the relationship among innovation, generative AI and non-linear thinking?” The answer is a playbook – which leads us back to how AI kills linear thinking.