More on this book
Community
Kindle Notes & Highlights
by
Azeem Azhar
According to complexity science, the connections between different elements mean that small changes in one area of a system can ripple across the whole.
One of the defining features of human history is our adaptability. When rapid technological change arrives, it first brings turmoil, then people adapt, and then eventually, we learn to thrive.
understanding new technology is hard. It takes knowledge of a wide range of new innovations. And it also takes an understanding of society’s existing rules, norms, institutions, and conventions. In
new technologies are being invented and scaled at an ever-faster pace, all while decreasing rapidly in price.
“exponential gap”: the chasm between new forms of technology—
and the corporations, employees, politics, and wider social norms that get left behind.
Technology is remaking politics, and politics is shaping technology. Any
was Microsoft and Intel that emerged from the evolutionary death match of the 1980s as the fittest of their respective species: the operating system and the central processing unit.
the computer of 1991 was millions of times more capable than that of 1981.
between 1971 and 2015, the number of transistors per chip multiplied nearly ten million times.
IBM for $150 apiece.14 By the 1960s, the price had fallen to $8 or so per transistor. By 1972, the year of my birth, the average cost of a transistor had fallen to fifteen cents,
The cost of a transistor had dropped to a few billionths of a dollar.
Put in the simplest terms, an exponential increase is anything that goes up by a constant proportion. A linear process is what happens to your age, increasing by a predictable one with each revolution of the earth around the sun. An exponential process is like a savings account with compound interest.
define an exponential technology as one that can improve at roughly a fixed cost, at a rate of more than 10 percent per year for several decades.
the computer chip business—with its roughly 50 percent annual improvement over five decades—
There are two sides to this phenomenon: decreases in price, and increases in potential.
When technologies develop exponentially, they lead to continually cheaper products which are able to do genuinely new things.
the true legacy of Moore’s Law. The hardware that underpins digital technology lends itself to continual increases in power and continual decreases in price.
While the paradigm described by Moore’s Law is reaching its limits, we are not reaching the limits of computational power in general.
It’s just that, in the future, increased computing power might not be based on cramming more transistors onto a chip.
computer can be considered intelligent if it is able to take actions that can achieve its objectives.28 Crucially, a piece of AI software needs to be able to make some kind of decision—rather than just blindly following each step in its program code.
computer scientist, John McCarthy,
Between 2012 and 2018, the amount of computer power used to train the largest AI models increased about six times faster than the rate of Moore’s Law.
you could fit about three thousand of them across the width of single strand of human hair.
The specialist AI chips from Graphcore are mapped at sixteen nanometers—more like 1,300 per strand.
computing power looks set to grow exponentially for the f...
This highlight has been truncated due to consecutive passage length restrictions.
waiting in the wings is a completely novel approach—“quantum computing.”
Classical computers would take hundreds of thousands of years to model such molecules; a quantum computer would take about a day.31
This primordial quantum computer was more than one billion times faster than its classical rival.
In 1975, a silicon photovoltaic module—
100 per watt of power it could produce.
Between 1975 and 2019, photovoltaics dropped in price some five hundred times—to under twenty-three cents per watt of power.
By October 2020, the cost of generating electricity from large-scale solar power, as well as from wind, had dropped below the cheapest form of fossil fuel product:
the ten years to 2019, the cost of generating electricity from wind turbines declined by 70 percent, or about 13 percent per year. This
The genome of a human is a long chain of binary information, some three billion letters long.
So to decode that first genome cost at least $500 million, and possibly as much as $1 billion.
August 2019, that price had declined to $942, a one-hundred-thousand-fold improvement.
That equates to a halving in price every year for two decades.
these innovations interact, they transform numerous parts of the economy. The smartphone has replaced many other consumer devices—Walkmans, calculators, diaries, watches, and street maps. Camera sales have collapsed in the wake of the rise of phone cameras.
from 2010 to 2020, the price of an average battery pack fell by nearly 90 percent.
2000: $50 bought you eight megabytes of storage.58 Twenty years later, $50 could buy you a good-quality two-terabyte flash drive (storing 250,000 times more of your memories). This represents an annual increase of 85 percent.
Exponential technologies are being driven by three mutually reinforcing factors—the transformative power of learning by doing, the increasing interaction and combination of new technologies, and the emergence of new networks of information and trade.
the tipping point at which exponential technologies began to really transform everything came in the second decade of the twenty-first century. In 2010, three hundred million smartphones were sold; by 2015, annual sales reached 1.5 billion.
Exxon was the world’s most valuable firm for the last time in 2013. By the beginning of 2016, six firms based on exponentially developing digital technologies—Apple, Tencent, Alphabet, Microsoft, Amazon, and Facebook—were among the top ten largest on earth.
Amazon doubled down on its investments in research. In the words of Werner Vogels, the firm’s chief technology officer, if they stopped innovating they “would be out of business in ten to fifteen years.”84
linear thinking, rooted in the assumption that change takes decades and not months, may have worked in the past—but not anymore.
every ten years, the cost of the processing that can be done by a computer will decline by a factor of one hundred.
As companies expanded and became more complex, managing them became like sorting through a mess of tangled wires—with too many contradictions resulting in slower and slower decision-making.
About 10 percent of the world’s public companies create four-fifths of all company profits.
“In 1980 it took twenty-five jobs to generate $1 million in manufacturing output in the US.” By 2016, it would take just five workers to produce the same value of finished goods.
