This is from my keynote address at MODSIM World Canada, delivered in June of 2010:
So what do soft drinks and exponential technology growth have to do with one another?
From John Sculley we know that however much you can persuade people to buy, that’s how much they’ll consume. This rule was formed around snack foods, but given how much we all use the Internet, my strong suspicion is that it applies equally to computing and communications. Let’s call this “Sculley’s Law”.
From Gordon Moore and Carver Mead, we know that computing power is doubling every 18 months, which equates to the price-performance of computing doubling in the same amount of time. This is Moore’s Law, but for the moment, let’s be generous to Dr. Mead and call it the “Moore-Mead Law”.
From Ray Kurzweil, we know that the Moore-Mead Law extends back to the beginning of the 20th Century, offering powerful historical evidence that exponential growth in computing power can survive technological paradigm shifts. This is an aspect of what is known as “Kurzweil’s Law of Accelerating Returns”, which tells us that in certain domains — specifically biology and technology — evolutionary processes tend to accelerate the pace of innovation.
From George Gilder, we know that total telecommunications bandwidth will triple every year for at least the next decade. This is “Gilder’s Law of the Telecosm”. Let’s simplify things and include in Gilder’s Law the related point that telecommunications bandwidth over any specific medium will double on the same time scale as the Moore-Mead Law.
And from Chris Anderson, we know that as the price of a commodity approaches zero, it becomes, in his words, “too cheap to matter”.
Do you see where all this is going? Computing and communications show every sign of continuing to increase in performance and decrease in cost at exponential rates for the foreseeable future. A single cycle of a CPU or a single bit of data delivered costs a hundredth of what it did a decade ago and a ten-thousandth of what it did two decades ago. And whatever we buy, that’s how much we consume.
However much faster Intel and other microprocessor vendors make their chips, we’ll use every cycle they give us. However faster telecommunications vendors make their networks, we’ll use every bit they give us. And they’re going to keep giving us more and more. We need a new law that sums up all of this. How about this: “Generally speaking, as the prices of a consumer commodity approaches zero, usage approaches infinity.”
Of course, that’s Economics 101. Price at zero consumption is infinity, and consumption at zero price is infinity. We need to qualify our law slightly. After all, do we expect that Pepsi could be made available in exponentially increasing amounts? Do we expect that if Pepsi were to lower its price to zero that people would consume infinite amounts of it? In 2055, the world consumed almost half a trillion liters of soft drinks. Were that to double every 18 months, in less than two decades, we’d be drinking the entire volume of Lake Ontario in soft drinks every year. Obviously there are limits to the production and consumption of tangible goods.
So we’ll modify our law slightly. Let’s say this: “For a unit of any given intangible commodity, over time, its price tends to approach zero and its usage tends to approach infinity.” I’ll call this “Boosman’s Law of Accelerating Usage”.
You can find the complete keynote address, including the reasoning behind each of the points made above, here.
Reading this text fresh, two years later, it seems to me to hold up well. I’d change a word or two here or there, but the basic conclusion still feels right.
The obvious question posed by the law I’ve hypothesized is what we’ll do with all the computing power and bandwidth we’ll have? If, 10 years from now, our computers and our connections to the Internet, both wired and wireless, will be 2 to the 6.67 faster, or 101.59 times faster, what will we do with that power and bandwidth? I’ll comment on that in future posts, or you can get an idea of where I’m going by reading the original address.