The Myth of Linear Progress
Why the Fallacy of Linear Thinking Is the Greatest Barrier to Investment Success
Introduction
I don't consider myself a conservative or a liberal. I am neither a value investor nor a growth investor. You might well describe me as a contrarian - someone who is known for questioning everything.
My intellectual curiosity led me to follow a wide range of thought leaders, investors, and decision makers across the political spectrum and philosophical worldviews. Over the past year, I have come to realize that most people have one thing in common: they fall victim to the fallacy of linear thinking. This blind spot affects leaders, investors, and decision-makers across the ideological spectrum.
Consider the polarized debate over renewable energy and electric vehicles:
Conservatives fixate on the limitations of today's solar panel and battery technologies - their material intensity, recycling challenges, range limitations, even the negative impact on the local landscape. They argue that today's fossil fuel technologies are superior in many ways. And arguably they are right, but only if we look at a very restrictive snapshot of today.
On the other side, progressives paint a bleak picture of unstoppable climate catastrophe and advocate drastic government intervention to force a transition to what they envision as a "sustainable green" future, which includes a vision of economic degrowth and other radical green-socialist policies.
Both camps fail to grasp a fundamental truth: the solution lies not in ideological dogma, but in the exponential advancement of technology itself. When it comes to the future, both sides make the false assumption that progress will mysteriously stagnate.
The reality is that solar photovoltaic technology is on an exponential trajectory and will – sooner rather than later – approach the theoretical limits. Battery energy density is on a similar exponential curve, poised to unlock transportation ranges measured not in miles, but in thousands of miles. Technological convergence will enable a decentralized, self-sustaining energy model unimaginable through a linear lens.
These dynamics go far beyond energy. Exponential advances in fields such as artificial intelligence, biotechnology, and materials science are radically reshaping our world in ways that linear thinkers cannot comprehend. Those wedded to linear thinking will continue to underestimate the scope and pace of change until the exponential returns render their worldview obsolete.
Overcoming our exponential blind spot is essential for leaders and investors to make informed decisions in the midst of this whirlwind of technological transformation. It requires a fundamental rewiring of how we perceive technological progress. Only by adopting an exponential view of how technology develops can we profit from the staggering changes on the horizon.
The Fallacy of the Intuitive Linear View
Most of us instinctively think about the future in terms of linear progress. When we envision what the world might look like in 10 or 20 years, we typically take today’s reality and project it forward at a steady, incremental rate of change. This linear way of thinking makes sense to us because it reflects our human experience of life unfolding in a sequential, straightforward manner.
However, this linear view causes us to grossly underestimate the potential of accelerating technological change driven by the exponential growth of innovations like computing power, artificial intelligence, renewable energy, biotechnology, and others.
Time and time again, we have seen expert predictions about the future proven laughably wrong because the forecasters failed to account for the exponential curve of progress (more on that in the next issue).
In 1977, Ken Olsen, MIT engineer and co-founder of Digital Equipment Corporation, famously said, “there is no reason for any individual to have a computer in his home”. Today, 4.8 billion people carry a smartphone in their pocket that has more processing power than the supercomputers of the 1990s.
When progress is exponential, straight-line forecasts break down completely. While linear growth proceeds in steady, incremental steps, exponential growth follows a radically different trajectory of accelerating returns. This divergence is deceptive at first, as the exponential and linear curves appear virtually identical in their early stages. But it is precisely this similarity that creates the “intuitive linear” mindset.
Humans are hardwired to perceive the future as a linear continuation of recent progress. We experience the world and the passage of time in an inherently linear fashion. Events unfold sequentially – one moment following the next in a straight chronological line. Our lived reality leads us to expect that future changes will also progress at a steady, predictable rate. We anchor ourselves in our present circumstances and simply extend that line forward. Furthermore, exponential growth involves counterintuitive concepts such as continual doubling over short periods of time. This simply defies our innate sense of numbers and feels implausible.
Example: If you take 25 steps of one meter each, you travel 25 meters – a reasonable linear projection. But if those steps double in length exponentially, your 25th stride would astonishingly span over 33 million meters, nearly the circumference of the Earth.
Again, when most people think about a future period, they intuitively assume that the current rate of progress will continue for future periods. This is what Ray Kurzweil calls the “intuitive linear” view. However, the intuitive linear view is wrong when applied to exponential systems and technologies. They fail because the rate of progress itself is not a constant, but doubles itself for each unit of time.
Relying on linear thinking in an exponential world sets us up to be repeatedly blindsided by unforeseen disruptions and to underestimate the full transformative potential on the horizon. As Ray Kurzweil says: We tend to overestimate what can be accomplished in a year and underestimate what can be achieved in 10 years. Exponentials make the future seem deceptively far away, until it arrives incredibly rapidly.
Ray Kurzweil points out that while we might intuitively expect 100 years of progress this century based on a linear view; the “historical exponential view” actually implies we will witness the equivalent of 20,000 years of progress by 2100 compared to the year 2000. This divergence is already playing out in the realm of computing and artificial intelligence.
Moore’s Law projected that transistor density on microchips would double every two years, and this exponential curve accurately described five decades of staggering progress. But AI capabilities are currently doubling at an even more explosive rate of every six months – four times faster than Moore’s Law. No other field in history has achieved such blistering exponential growth.
The exponential growth of AI itself will catalyze further technological breakthroughs – for example in solar and battery technologies – through intelligent system optimization, accelerated materials discovery, and predictive simulation capabilities (but more on that later).
Calculating Exponential Growth
Let's make a mathematical excursion and calculate the factor by which the capacity of AI technology will increase, assuming exponential growth with the current doubling every 6 months.
For this, we use the formula for exponential growth:
Given that the current doubling period of AI is 6 months, we first need to convert years into periods of 6 months:
5 years = 10 periods (since 5 years * 2 periods per year)
10 years = 20 periods
25 years = 50 periods
50 years = 100 periods
Now, applying the formula:
For 5 years:
For 10 years:
For 25 years:
For 50 years:
Therefore, under the assumption of exponential growth with a doubling every 6 months, the capacity of artificial intelligence technology will increase by a factor of 1,024 in 5 years, approximately 1.05 million in 10 years, approximately 1.13 quadrillion in 25 years, and approximately 1.27 sextillion in 50 years.
In other words, given the current improvements of AI (doubling every 6 months), and projecting it on an exponential scale will result in the year 2074 in technology that is 1,267,650,600,000,000,000,000,000,000,000x more advanced than the AI we have today.
Sextillion is a number so large that you have probably never heard of it. This factor is too large for us to imagine, which makes it impossible and impractical for us to work with it.
Real-World Examples
While the concept of exponential growth may still seem like an abstract mathematical concept, we are surrounded by powerful examples of technologies undergoing exponential progressions. From computing power and internet adoption to renewable energy, biotechnology, and artificial intelligence, the pace of progress is accelerating at an exponential rate (not linearly).
Let’s look at some real-world examples that illustrate the exponential trajectory of technology.
Moore’s Law and Computing Power
Perhaps the best-known and most cited example of exponential growth is Moore’s Law, which predicts that the number of transistors on a computer chip will double approximately every two years. This exponential increase in transistor density has driven the staggering growth in computing power over the past five decades.
In 1965, Intel co-founder Gordon Moore observed that the number of components on a chip was doubling every year. He predicted that this trend would continue. In 1975, he revised the doubling period to every two years and predicted that this trend would continue. Remarkably, this prediction has held true, with the number of transistors growing from a mere 3,500 in the Intel 8008 chip in 1972 to nearly 50 billion in modern processors.
The exponential growth in computing power itself has been the driving force behind other technologies of the digital revolution, enabling the development of increasingly sophisticated software, the internet, and artificial intelligence.
DNA Sequencing
The cost of sequencing a human genome has dropped exponentially from $100 million in 2001 to about $100 today. This is enabling personalized medicine, gene editing tools like CRISPR, biohacking, and even the potential revival of extinct species. But the exponential curve of biotechnology’s capabilities has only just begun!
Renewable Energy
The costs of solar photovoltaics and battery storage have followed exponentially declining cost curves. This is igniting a rapid disruption of fossil fuels and exponential growth of renewable energy generation capacity. Exponential energy deflation – that means at one point energy prices too low to meter – will power many other exponential technologies that have high energy requirements (such as artificial intelligence).
Artificial Intelligence
The performance of AI systems, particularly in benchmark tests like image recognition and natural language processing, has shown remarkable improvement over recent years. While the rate of capability enhancement varies, advancements in data availability, computing power, and algorithmic efficiency have collectively propelled AI forward. Although the exact rate of doubling in AI capabilities is subject to debate and varies by application, the overall trajectory points to AI as an exponentially improving technology and a transformative force for any other technology.
AI: A General-Purpose Technology
In order to filly grasp the exponential technological progress ahead of us, it is crucial to understand the transformative power of artificial intelligence as a general-purpose technology. Unlike useful discoveries – such as dental floss or the polio vaccine – a general-purpose technology – such as the steam engine, electricity, or the internet – is not useful for one specific application only but can be applied to numerous purposes.
General-purpose technologies have not merely served a single purpose, but have reshaped entire economies, societies, and the very fabric of human civilization. The steam engine ushered in the Industrial Revolution, electricity powered the modern age of manufacturing and communication, and the internet gave rise to the digital era, spawning countless new business models and industries.
Artificial intelligence is poised to be the next general-purpose technology, with implications that will dwarf those of its predecessors. Its versatility stems for the ability to process vast amounts of data, identify patterns, and optimize complex systems – capabilities applicable to virtually every field of human endeavor.
But AI’s true disruptive potential lies in its ability to accelerate the exponential progress of other technologies. As a meta-technology, AI can supercharge innovation cycles by augmenting scientific discovery, optimizing design processes, and even automating aspects of research and development itself.
This positive feedback loop exemplifies the law of accelerating returns described by Ray Kurzweil; whereby technological progress begets more progress in a self-reinforcing cycle.
For example, AI is already revolutionizing biotechnology by rapidly screening millions of molecular compounds, optimizing protein design, and even aiding in the development of novel gene therapies. These biotech breakthroughs could then enable engineered organisms to produce advanced nanomaterials and biological computers, further enhancing the hardware capabilities of AI. More powerful AI systems, in turn, could unlock abundant renewable energy by optimizing solar cell designs and grid management - and provide the massive computing power needed to train the next generation of AI models.
This infinite cycle extends across domains, with AI driving advances in materials science, energy storage, transportation, and even the reverse engineering of the human brain. Each AI-enabled breakthrough generates tools and insights that feed back into the system, catalyzing further exponentials in a continuous cycle of reinvention.
Recognizing AI's role as a general-purpose technology is critical for investors, entrepreneurs, and policymakers alike. Just as previous general-purpose technologies made entire industries obsolete while giving rise to new ones, AI will inevitably disrupt and reshape the economic landscape in ways we cannot yet imagine.
The Six Ds of Exponentials
While the exponential growth patterns of transformative technologies such as AI are powerful drivers of change, their implications can be challenging to grasp intuitively. To better understand how exponential technologies such as artificial intelligence develop, Peter Diamandis divides the exponential growth cycle into six key steps, which he named the “Six Ds of Exponentials” that characterize their evolution and diffusion: digitization, deception, disruption, demonetization, dematerialization, and democratization.
In the first phase, digitized, anything that is digital or anything that can be digitized has the ability to enter an exponential growth phase. This includes AI, which I estimate has currently already entered the second phase: deception. At the beginning, exponential technologies don’t seem to grow very fast – which is deceptive. 0,1 only becomes 0,2. Yet as we break the whole-number barrier, exponential growth accelerates faster: 2 quickly becomes 32, which becomes 64,000 faster than we can grasp. Do you remember the previous example of making 25 linear versus exponential steps?
In the third phase, a digitalized, exponentially growing technology will disrupt existing markets and industries by outperforming them in effectiveness and cost. Once you can use Excel, why manually crunch numbers?
The fourth phase, demonetization, states that exponential technologies become cheaper, often to the point of being free. The barriers to entry for adopting AI-driven solutions are rapidly diminishing. For example, Chat GPT-4 is currently available for free or at a very low cost. In the future, much more advanced models will become available for free or close to free.
In the fifth phase, dematerialization, physical products disappear. Today, a smartphone combines hundreds of technologies which were once bulky and expensive: radio, music player, camera, GPS, video calls, maps, encyclopedias, etc. AI will lead to a similar dematerialization; this will be in the form of jobs but also entire supply chains and manufacturing processes through the emergence of highly advanced 3D printing technologies and much more – which all work synergistically with AI.
In the sixth and last phase, democratization, once something undergoes digitization, its accessibility broadens. Consider the fact that the wealthiest and most influential individuals in the world use the same smartphones as the average person. Consider that equally, today the wealthiest and poorest persons have the same access to the most advanced generative AI through services like ChatGPT and open-source models Llama 2.
In the case of AI as a general-purpose technology, we are likely still in the deceptive phase, with its full disruptive potential yet to be unleashed.
Overcoming the Intuitive Linear Bias
Despite the mounting evidence of exponential technological change unfolding right in front of us, our deeply ingrained bias for intuitive linear thinking blinds us to understand the true magnitude of technological progress. Overcoming these cognitive hurdles is perhaps the greatest challenge we face in preparing for and harnessing the exponential age.
Again: the roots of our linear thinking run deep, shaped by millions of years of evolutionary experience in which change happened gradually. Our brains evolved to process information linearly, extrapolating the near-term future as a continuation of recent trends. This way of thinking served our ancestors well in a world of relatively stable, incremental change.
However, the exponential progression of technologies such as AI, biotechnology, and energy systems represents a radical departure from this linear norm.
The mismatch between our linear intuition and the historical exponential reality manifests itself in several cognitive biases that impede our ability to grasp the exponential change:
Exponential Growth Bias: The exponential growth bias is the tendency for people to underestimate the long-term effects of exponential growth and compounding rates of change. When we observe an exponential curve locally or over a short time horizon, it appears linear. This localized perspective causes us to linearize the curve and consistently underestimate the long-term impacts of compounding growth rates.
Magnitude Bias: The magnitude bias is the human tendency to struggle with comprehension and emotional resonance that exceeds the scale of human experience. The sheer magnitude of exponential growth quickly exceeds our cognitive grasp, making it difficult to impossible to fully grasp the implications of long-term exponential progress.
Status-Quo Bias: The status quo bias is the deep-rooted tendency of people to prefer and maintain the current status quo, even when presented with potentially better alternatives or when change is necessary. It is an irrational preference for the familiar and a resistance to change that leads us to simultaneously assume that the future will resemble the past and present, dismissing potential disruptions from the norm as aberrations.
Confirmation Bias: Confirmation bias is the tendency for people to selectively seek out and interpret information in a way that confirms our pre-existing beliefs and worldviews, leading us to disregard any evidence that challenges our linear assumptions.
The sum of these biases leads us to our intuitive linear bias, which is the cognitive illusion that leads us to expect the future to unfold at the same pace as the past.
Overcoming these biases requires a concentrated effort to rewire our thinking patterns to cultivate an “exponential mindset”. By regularly exposing ourselves to the principles of exponential thinking – for example by practicing looking at a particular technology through the lens of the “Six Ds of Exponentials”, reading on historical case studies of exponential growth, immersing ourselves in data that challenges our intuitions, and embracing the mindset of questioning long-held assumptions – we can recalibrate our expectations and prepare for the exponential pace of innovation.
Ultimately, transcending linear biases is a necessity for navigating the exponential age as a leader and investor. Those who fail to internalize an exponential mindset risk becoming obsolete as the pace of change accelerates.
While it is impossible to reliably predict the outcomes of any specific project or business, Ray Kurzweil emphasizes that, despite the apparent chaos, the overall progress of information technology is following an exponential path. This allows anyone who embraces an exponential worldwide to reliably anticipate the future and seize emerging opportunities that are invisible to the linear thinker.
Closing Words
The electric vehicle and solar energy debate from the introduction illustrates how both conservatives and progressives fall into the trap of linear thinking. Conservatives dismiss exponential improvements in solar PV and batteries. Progressives underestimate how quickly free-market entrepreneurs can make clean technology leapfrog fossil fuels and solve the world's biggest problems.
The truth is, the future isn't predetermined - it's built by entrepreneurs who bring exponential technologies to market. The dynamism of free market capitalism is driving exponential progress in solving grand challenges like climate change, disease, and energy shortages.
While doomsayers wring their hands, visionary capitalists are betting on exponential fields like AI, nanotech, and synbio to shape an abundant future - including 99.97% efficient solar panels to meet all energy needs.
The future belongs to those who think exponentially, not linearly. Investors, leaders and policymakers must embrace the exponential or be made obsolete by those who do. The myths of linearity are crumbling - the future is exponential.
Coming Next: The Forecasting Fallacy
You just read how linear thinking blinds us to the exponential potential of transformative technologies.
In the next edition, we’ll dive into the "forecasting fallacy" - the tendency for even the most renowned analysts, consultancies, and institutions to grossly miscalculate the timing and impact of technological change.
You'll see striking examples of how respected voices made laughably inaccurate predictions that now seem absurd in hindsight. From Paul Ehrlich's warnings of mass famine and societal collapse to McKinsey's projections that underestimated AI's disruption of the workforce by orders of magnitude, we'll deconstruct why the best and brightest consistently fail to anticipate the true exponential curve.
More importantly, you'll learn alternative frameworks for long-term forecasting that go beyond simplistic linear models. We'll explore approaches such as environmental futurism, Kondratiev wave analysis, generational cycles, geopolitical power shifts, and the study of technological revolutions.
Only by combining these multidisciplinary lenses can we develop an accurate picture of the path to AGI, the Singularity, and all the risks and opportunities that this journey presents to business leaders and investors.
Don’t forget Cancer.