# I’ve Spent 33 Years Watching Crazy Ideas Become Ordinary. Here’s What I’ve Learned.
By 

## Article Content
The greatest mistake in predicting the future isn’t getting the technology wrong. It’s underestimating how quickly the impossible becomes routine.A few days ago, I found myself on stage in Chicago, speaking to senior executives from the electronics industry about the extraordinary technologies reshaping their world. I was delivering the keynote for the Electronic Components Industry Association (ECIA) Executive Conference, exploring the collision of artificial intelligence, semiconductors, energy, robotics, geopolitics and accelerating technological change.At one point, I put up a slide about silicon photonics and the growing importance of light-based technologies in the massive global buildout of artificial intelligence infrastructure.And I had one of those moments that occasionally happen after more than three decades of working as a futurist. I realized I had been talking about this general technological direction for more than a quarter-century.In 2000, my co-author Rick Broadhead and I published From Light Bulbs to Yottabits: How to Profit by Understanding the Internet of the Future. We devoted an entire chapter to photons, optical technologies and the extraordinary possibilities that might emerge as light became an increasingly important part of digital networks. I even registered the domain YottaBit.com that year.In 2006, I stood before thousands of telecommunications engineers and challenged them to imagine future networks and devices that would use light in ways far beyond the engineering constraints of the day. Some laughed. A few complained to the conference organizers. Later, an industry publication questioned what I had suggested. In 2007, I wrote about the future of optical chips built around light rather than electrons. Now, in October 2026, artificial intelligence infrastructure is creating significant demand for optical interconnects and silicon photonics. That doesn’t mean every ambitious prediction I made has come true. Optical computing and optical data transmission are distinct, and many proposed breakthroughs are still research challenges. But the underlying direction has become an important part of the technology story. I’ve told that story in more detail in my article about AI, copper and light, and documented the original controversy in this retrospective from my archive. My recent ECIA keynote case study shows why this is not simply a telecommunications story anymore. But this article isn’t really about photonics. It’s about what 33 years of watching apparently crazy possibilities move toward practical reality has taught me about the way we misunderstand the future. Three points in a much longer story: early research, the challenge of communicating emerging possibilities, and the present-day infrastructure implications. 1. The future often sounds ridiculous before it becomes obvious In the early years of the commercial Internet, I spent enormous amounts of time helping organizations imagine what might happen when connectivity became pervasive, information became ubiquitous and entirely new business models emerged. There was skepticism about online finance, digital distribution, connected services and all manner of possibilities that would later become routine. Some of that skepticism was reasonable. The technology was immature, business models were uncertain and infrastructure was often inadequate. But another problem was at work: people were evaluating tomorrow’s possibilities using today’s capabilities, economics and customer expectations. They were confusing not practical today with will never be possible. A futurist’s job isn’t to declare every strange idea inevitable. Most won’t succeed. The useful work lies in identifying credible trajectories in science, technology, cost and adoption, and asking what might happen when multiple constraints change together. 2. Science fiction has an annoying habit of becoming ordinary For years, one of my favorite keynote openings has involved an old television show: The Jetsons. The series debuted in 1962 and imagined a world set in 2062. The specific details were often whimsical, but the broad imagery of video calls, automated household technology, intelligent devices and robotics has proven remarkably useful for showing audiences how quickly remarkable technology becomes familiar. We still don’t have flying cars in every driveway, and no cartoon predicts the future perfectly. That’s precisely the point. Think about how casually we now conduct a video call from a phone, ask an AI system to work through a complex problem or summon navigation instructions from an invisible network. The extraordinary becomes mundane almost as soon as it works reliably. Yesterday’s science fiction becomes today’s infrastructure. And today’s infrastructure becomes tomorrow’s assumption. We become so comfortable with what has arrived that we fail to notice what may be next. The familiar progression from impossible-sounding idea to everyday capability. 3. The biggest forecasting mistake is thinking in straight lines If something took ten years to develop in the past, we assume the next version will take another ten years. If a breakthrough is expensive today, we presume it will remain expensive. But falling costs, better tools, expanding knowledge and intersecting disciplines can accelerate change in ways that defeat conventional extrapolation. I’ve watched this pattern across computing, communications, genomics, artificial intelligence, industrial automation and energy technologies. Not every curve continues exponentially forever, and adoption is always constrained by regulation, capital, skills and human behavior. But linear assumptions can be spectacularly misleading when several underlying capabilities improve together. The question I encourage audiences to ask is not merely whether a technology works today. It is: what becomes possible when the capability improves, the price drops, the infrastructure matures and new participants enter the market? 4. Expertise is valuable. It can also make us blind I have enormous respect for engineers and executives who devote their careers to mastering complex systems. Their expertise makes modern society work. But expertise is often built around an existing set of assumptions. When those assumptions change, yesterday’s knowledge can become an anchor as easily as an advantage. Consider a recent private briefing I developed for senior executives from iA Financial Group – Automotive. Traditional warranty and protection businesses were built around mechanical failures, familiar repair patterns, dealership distribution and historical underwriting experience. What happens when vehicles become software-defined and connected, when risk shifts toward batteries and electronics, or when the manufacturer increasingly owns the customer relationship? Those aren’t simply technology questions. They are questions about the future of an entire business model. The more successful an organization has been under the old model, the harder it can be to recognize when that model requires reinvention. Your greatest competitive liability might be the assumptions that made you successful in the first place. Original iA Financial Group keynote case study image. Click to explore the customized briefing. 5. The future doesn’t arrive one trend at a time At ECIA in Chicago, AI was driving a discussion about semiconductors, but that was only the beginning. The expansion of AI infrastructure creates demand for electricity, optical connectivity, cooling and advanced manufacturing. Robotics changes component demand. Geopolitics changes sourcing. The energy transition changes the shape of industrial investment. These forces are not separate stories. They collide. This is why the research behind each of my keynotes must be particular to the audience. The same underlying technologies mean something different for a semiconductor distributor, an energy utility, a risk professional or an automotive warranty provider. The critical task is to translate broad patterns into specific implications for strategy, products, risk and leadership. A good keynote should not leave executives merely impressed by a dazzling list of trends. It should help them ask better questions about the decisions they must make when they return to work. 6. Being early only matters if you keep paying attention I did not write about photons in 2000, present a provocative scenario in 2006 and then forget the subject. The useful part of foresight is a willingness to revisit assumptions as evidence accumulates. Sometimes the technological direction is right but the timetable is wrong. Sometimes a breakthrough arrives in a completely unexpected form. Sometimes the expected commercial application never materializes. That is why I believe predictions should be updated rather than defended as monuments to the past. In 2026, I launched YottaBit.com, returning to the domain I registered in 2000 and using it as a home for exploring the scientific and technological possibilities that are emerging now. It reflects the same curiosity that drove our original book, informed by everything that has happened since. 7. “That will never happen” is a dangerous way to think I’ve heard it in just about every industry: customers will never behave that way; our distribution model won’t change; the regulatory system won’t permit it; our competitors would never do that; our business is different. Sometimes an objection is valid. But the danger begins when skepticism becomes a substitute for investigation. The better question is: what would have to change for this possibility to become practical? What if the cost falls? What if the science advances? What if a new customer expectation emerges? What if a competitor arrives without any investment in the old business model? Suddenly, a dismissal becomes a useful exercise in strategy. 8. The future rewards curiosity more than certainty I don’t believe the purpose of a futurist is to provide a perfect map of 2040. The future is shaped by too many interdependent forces. What three decades of experience can provide is perspective: an ability to identify patterns, challenge comfortable assumptions, recognize emerging signals and help people understand what might matter next. The greatest result of that work is not that everyone believes a particular prediction. It is that a leadership team leaves asking better questions. Which assumptions behind our strategy are already becoming obsolete? What might happen faster than we expect? What could emerge from outside our industry? And what should we begin testing today? The crazy becomes routine. Then something else becomes crazy. Looking back, the pattern is striking. An idea begins as something impractical, unrealistic or ridiculous. The underlying science advances. Costs decline. Early experimenters find workable applications. Supporting infrastructure appears. The idea becomes commercially viable and then ordinary. Almost immediately we begin dismissing the next wave of possibilities because they sound just as strange. That is the remarkable cycle I’ve spent 33 years watching. It is why I keep revisiting old slides and books, not simply to see where I was right, but to understand what the journey from possibility to reality can teach us about the next wave. A closing thought on the need to remain curious, adaptable and ready to rethink old assumptions. The future doesn’t care whether you are ready to believe in it. It just keeps arriving. And the ideas that sound crazy today? Some of them are already on their way to becoming tomorrow’s ordinary reality. The question is whether we will recognize them early enough to do something about it. For further reading: the optical prediction retrospective; the 2026 ECIA keynote case study; the YottaBit research project. %

---
Source: https://jimcarroll.com/2026/10/ive-spent-33-years-watching-crazy-ideas-become-ordinary-heres-what-ive-learned/