🤖 Effective Accelerationism (aka e/acc)
🤖 Effective Accelerationism (aka e/acc)
The rate of adoption of revolutionary technology increases year over year
Historically, new technology takes time to make an impact. In The New American Dream, I compared large-language models to the power loom, a machine that completely revolutionized the creation of textiles during the industrial revolution.
Despite its massive impact on productivity, it took fifty years to reach widespread adoption. One of the reasons for this slow adoption was a lagging rate of production. They had mechanized the looming process but had not mechanized the creation of the loom.
Let's compare the power loom to the adoption of the automobile. Automobiles have greatly influenced the world and arguably are the invention that brought us into the modern age. They were introduced to Americans in the early 20th century by Henry Ford. In the chart below, you can see that cars had a much higher adoption rate than the power loom, overtaking horses in less than 25 years. Ford's Model T overcame the supply constraints the power loom faced by revolutionizing the assembly line, scaling up the production of cars to meet the demand of millions of Americans.
The adoption of revolutionary technology is limited by its capacity for distribution. As humanity increases our ability to share information, produce tools, and collaborate effectively, the speed at which we can adopt and adapt to novel technologies is growing exponentially. This trend can easily be seen below in this chart from the Harvard Business Review. As time passes, the rate of adoption for new technology increases.
Today, we are standing at the precipice of another technological revolution – the AI revolution. AI models, especially large language models like GPT-4, represent a radical shift in our approach to problem-solving, innovation, and productivity. They are capable of understanding and generating human-like text, opening doors to numerous applications across various industries, from writing and editing to customer service and even scientific research.
Much like the power loom and the automobile, the reach and influence of AI models hinge on their distribution and adoption. But here's where AI has an advantage - it is software, which allows for far more rapid and broad-scale distribution compared to any physical technology because it isn't constrained by physical production or logistics. Once a new AI model is developed, it can be distributed globally in mere seconds at virtually no additional cost. And as we know, software is eating the world.
But rapid distribution isn’t the only factor contributing to AI’s potential for swift and widespread adoption. Modern software development practices, along with a sophisticated array of developer tools and APIs, have made integrating AI into existing software and systems easier than ever. This not only makes AI more accessible to a wide range of industries and disciplines but also helps in extending existing use cases, further propelling its adoption.
In conclusion, the essence of our technological journey is that each innovation builds on its predecessor, paving the way for faster adoption of future breakthroughs. As Ray Kurzweil, the renowned futurist and advocate of the theory of accelerating returns, eloquently puts it, "We won't experience 100 years of progress in the 21st century—it will be more like 20,000 years of progress."
The assembly line facilitated the quick adoption of automobiles, computers led to the internet, and now, AI is becoming a foundational layer of technology for automated problem-solving. Each technological advance effectively sets the stage for the next, accelerating the pace of innovation and adoption.
The future is poised to adopt new technologies at a pace that eclipses anything we've experienced so far, promising a transformative era of accelerated growth and innovation. This, indeed, is an exciting time to be alive. In Kurzweil's terms, we may be gearing up for not just decades, but millennia of progress, all within our lifetimes.
George and I are currently in Canada for our first Hawk Hill Ventures offsite. We’ve rode 175 miles over the past six days, crossing through Glacier National Park, the Canadian Border, and Banff National Park.
