Use the AI S-curve to drive meaningful technological change

In the early 20th century, the Wright brothers’ first flight barely left the ground, lasting just 12 seconds. But within a few decades, airplanes revolutionized global travel and connected the world in ways previously unimaginable. Today, a similar pattern is unfolding with artificial intelligence as it transitions from a niche innovation to a ubiquitous tool that is reshaping industries worldwide. Technical revolutions like this can be visualized and understood using a framework called the “S-curve.”

The S-curve is a graphical representation of how technology matures over time. It starts slowly, with early adopters, specialized use cases, and technocrats. As the technology proves its value, it enters a phase of rapid growth where adoption accelerates and becomes more widely integrated into various industries and applications.

However, as technology advances, becoming cheaper, faster, and more efficient, it inevitably reaches some logical limit (often defined by some practical physical limitation based on the laws of physics) and settles into a natural “top” of the S-curve. When a technology reaches its limit, progress is relatively slow, typically requiring significant increases in complexity. For example, look at efficiency gains by the internal combustion engine over the last 20 years. Over time, a new technology emerges, typically starting at a lower performance level than the original, but it results in a new S-curve that has the potential to overtake the old one.

Understanding this lumpy pattern of technological evolution is essential for organizations that want to make informed decisions about when to invest in and adopt new technologies today and in the future. This will be particularly true for AI-based or enabled technologies.

What’s driving the pace of the AI S-curve?

AI does not exist in a vacuum; it is part of a broader ecosystem of technologies that enable use cases. To truly understand the trajectory of AI adoption (or any technology), it’s important to look at the synergies between other technologies. For example, transformers, a type of neural network architecture, have revolutionized the way AI models process and generate human language; we’ve all heard of large language models (LLMs). However, transformers alone are not the only technology creating excitement around LLMs like GPT-3. Rather, it’s the transformer applied to LLMs and a bunch of other enabling technology that unlocks new possibilities for AI.

The exponential growth in computing power, as described by Moore’s Law, has been—and continues to be—a fundamental driver of the AI technology S-curve. This ever-increasing computational capacity allows AI models to process larger and larger data sets, handle more complex algorithms, and find solutions to problems previously considered intractable. The point is that organizations must account for enabling technologies, such as cloud computing and big data analytics, Moore’s Law, and the open-source software movement, as well as work in synergy with research in the field of AI itself, to further accelerate the pace of AI innovation and adoption.

Disruptions are nonlinear and can come from outside the core field. We may see the next wave of AI driven by completely new forms of compute. For example, neuromorphic computers with in-memory processing are non-Von Neumann systems that, if perfected, would greatly reduce the amount of compute power needed to perform matrix mathematics, offering tremendous benefits for AI model execution.

Separating the signal from the noise

Today’s AI landscape is a lot like a Jackson Pollock painting—a seemingly chaotic array of possibilities and experiments. But just like Pollock’s paintings were determined to be based on fractals, there are patterns at the root of progress in the AI space. As AI advances, it’s easy to get caught up in the hype and apply it to every problem we encounter.

It’s important to recognize that some seemingly new AI applications are not actually solving new problems; they are simply applying AI to tasks we already know how to solve through other means. It’s like the adage, “When you’re given a hammer, everything looks like a nail.”

The real challenge (and value) lies in separating the signal from the noise—identifying the areas where AI can truly make an impact and provide novel solutions to complex problems.

Balancing risk and reward in the age of AI

As new technologies like AI emerge and mature, organizations must balance the need to stay competitive with the potential risks and uncertainties associated with early adoption.

This challenge is not new. In his book, The Innovator’s Dilemma, Clayton Christensen describes the difficult choice companies face between maintaining their existing, profitable business models and investing in new, potentially disruptive technologies. So, how can organizations navigate this decision?

One approach is to ensure that there is a dedicated unit that operates on a long takt time, outside the quarterly or annual reporting pressure. By having a unit that operates on long cycles with a specific focus on deep technology and new business models, organizations can create an “early warning radar” for disruptive changes and trends. This unit’s job is to scan the horizon, identify potential disruptions, validate them, and raise these issues to senior levels of the organization for synthesis. They focus on questions such as:

What does this new technology mean for our business?

Is it a potential disruption, an opportunity, or both?

How do we need to adapt our strategies and capabilities in response?

Sometimes, this unit must explain that the current technology is getting to the top of an S-curve and that additional gains with the same approach may not be worth the incremental investment, even when the disruptive technology is less performant than the incumbent. 

Three horizons of AI-fueled growth

For organizations with limited resources, a different approach is needed. It’s crucial to first understand what mode you’re operating in. Are you in a scientific mode, seeking to understand the underlying principles of a technology? Or are you in an engineering mode, focused on applying the technology to solve specific problems?

If you don’t have the resources to separate these modes, it’s important to devote your time and attention to three different horizons:

The first horizon is the present. What’s happening today, and how can we optimize our current operations?

The second horizon is the near future. What’s coming next, and how can we position ourselves to capitalize on those opportunities?

The third horizon is the far future. What’s out there on the distant horizon, and how might it shape our industry in the long run?

By monitoring updates and trends across these horizons, organizations can stay attuned to the nonlinear development of emerging technologies and limit the potential for strategic surprise to the business. This doesn’t necessarily require a stand-alone unit, but it does require a stand-alone activity—a deliberate effort to step back from the day-to-day and consider the bigger picture.

The art of anticipation

As we navigate the rapidly evolving AI landscape, we need to approach it with both a wondrous and a wary eye. The S-curve is a reminder that the journey ahead is not a linear path, but rather a dynamic and iterative process. By staying attuned to the dynamics of the AI S-curve and the factors influencing adoption, organizations can strategically position themselves for success.

Kyle Crum is director of advanced technology at Rockwell Automation.

—

Generative AI Insights provides a venue for technology leaders—including vendors and other outside contributors—to explore and discuss the challenges and opportunities of generative artificial intelligence. The selection is wide-ranging, from technology deep dives to case studies to expert opinion, but also subjective, based on our judgment of which topics and treatments will best serve InfoWorld’s technically sophisticated audience. InfoWorld does not accept marketing collateral for publication and reserves the right to edit all contributed content. Contact doug_dineley@foundryco.com.