Kevin Kelly: "12 Laws That Will Shape the Future" (Part 1)

The Japanese translation of Kevin Kelly's new book, " What Comes After the Internet: 12 Laws That Will Shape the Future," was released in Japan in late July. Kelly, founding editor of WIRED magazine and a leading thinker in the technology world, explained three major trends emerging from the significant changes expected over the next 30 years during a commemorative lecture for the book's publication at this Dentsu Design Talk.

 

Can the Future Be Predicted?

Today, I'd like to talk about the future 30 years from now.
What direction will the future take? Much of it remains unpredictable, an unknown territory. However, some aspects can be forecasted. Today, I will discuss these foreseeable directions.

First, we must understand that technology is not a "single entity." It is not just the individual microphone or computer I am using now, but rather "one large system." Various things are interconnected, forming a vast network, which I call the "Technosphere."
Even when technologies exist independently, they exhibit recurring patterns, with vectors tilting toward specific directions. Where does this tilt originate? It stems from the very nature of the technologies that make up the system.

Physical entities like electrical switches, silicon chips, or wiring determine the overall inclination—the directionality. This is the "technosphere," setting the long-term course. Thus, by observing this inclination of technology, long-term trends become predictable.

Consider, for example, the "four legs." Four legs are a highly stable configuration when considering the physical laws of gravity. Therefore, on any planet with gravity, animals would likely walk on four legs, and vehicles would have four wheels. This is determined by the physical laws of the real world.
In other words, the form of four legs itself is inevitable, but the existence of individual species is not inevitable. This means that while the existence of four-legged animals is inevitable, whether they become a specific species like the zebra is not inevitable.

Let's consider another example. Imagine rain pouring down into a valley. The path each individual raindrop takes to reach the valley floor is random and unpredictable. However, the direction "flowing downward" is inevitable. This is because the force of gravity moves the water in a specific direction.

So, what about the realm of technology and digital systems?
If we think broadly, the form of the "telephone" is inevitable. Given electricity and wiring, the technology of the telephone will emerge regardless of culture, era, or political system. On the other hand, the iPhone is not inevitable. We cannot predict individual products or companies that will emerge.

Like the telephone, the "internet" is also an inevitable presence. Once the telephone existed, its evolution into the internet was unavoidable. However, whether Twitter would emerge from that is not inevitable.

From this broad perspective, let's look ahead to the future 20 or 30 years from now. It resembles a vast river with intricate tributaries. Each tributary is independent yet simultaneously interdependent. I view all 12 tributaries I listed in my new book, What Comes After the Internet, as ongoing processes—things that are unfolding.
Each trend already exists today and will intensify in the future. Today, I'll explain three crucial trends among them.

 

The most important direction is "COGNIFYING"

The most important trend is "COGNIFYING" (the process of becoming cognitive). This is a verb, a coined term meaning "to make things smarter."
AI (artificial intelligence) already exists in the modern world. It has been slowly evolving over the past 50 years. Examples like iPhone's "Siri" or Amazon's Echo are easy to understand, but most AI remains invisible in daily life.

For instance, in hospitals, AI diagnoses X-ray results because it can do so more accurately than doctors. Or in law firms, AI is used to scrutinize evidence because it's more efficient than human lawyers. AI is also used for autopilot in airplanes. Even now, AI handles most of the flying, and the time spent by human pilots is only a fraction of the total flight time. Modern cars also have AI chips in their brakes, operating them more effectively than humans. While all these exist already, we don't typically see them in our daily lives.

And over the past five years, three new technologies have been integrated with this AI, driving its advancement.
The first of these technologies is software called "neural networks." While the algorithm was developed 50 years ago, scaling it up successfully had previously been difficult. Recently, Canadian researchers discovered that stacking neural networks in layers could achieve this. This is called "deep learning."

The second technology relates to hardware. Previously, implementing AI required extremely expensive supercomputers costing millions of dollars. However, about five years ago, small chips used in video games and similar applications became affordable. This made GPUs (Graphics Processing Units) a new platform capable of running AI.

The third technology concerns big data processing capabilities. Training AI requires massive amounts of data; it must learn from thousands or even millions of examples. The only way to prepare that much data was not to create it from scratch, but to extract it from everyday data. Companies like Amazon, Google, Baidu, or Microsoft accumulate and utilize vast amounts of data during searches.
With these three technologies—deep learning, GPUs, and big data—all in place, truly functional AI became achievable.

 

Understanding the Difference Between AI and Human Intelligence

Google's "AlphaGo" defeated the world's top Go player. By utilizing big data, running on GPUs, and employing deep neural networks, this AI was able to play creative moves and defeat humans.
While there are various prejudices against artificial intelligence, this kind of intelligence is truly fascinating. For example, computers are smarter than you when it comes to calculations. GPS is smarter than you for spatial navigation. So, we utilize AI that is smarter than humans in these specific domains.

For instance, imagine equipping a car with AI. Unlike humans, AI doesn't get distracted. It won't wonder while driving, "Did I turn off the stove back home?" or "Maybe I should have majored in something else at university." Consequently, the selling point for AI-equipped cars becomes "it has no consciousness." That's preferable.

We tend to think of intelligence as one-dimensional, improving in a linear fashion. For example, mice have the lowest IQ among animals, chimpanzees have a slightly higher one, and humans have the highest. And among humans, there are those with low IQs, average ones, and geniuses. But this is a mistaken view of intelligence.

Our brains and intelligence are complex, combining dozens or even hundreds of different types of thinking. For example, human intelligence combines various types of thinking like deductive reasoning, emotional intelligence, and long-term memory. It's more like a symphony, with various instruments playing their parts.

Animals also possess different symphonies. Some animals play instruments similar to ours, while others demonstrate abilities superior to humans in specific domains. For example, squirrels can remember where they buried nuts years ago, possessing memory capabilities superior to humans.
Many AIs enhance these multiple cognitive abilities. Crucially, they possess a different kind of thinking than humans. I believe this represents a Copernican revolution.

We tend to assume human intelligence is universal. That is, we believe our intelligence is always superior to the intelligence around us. And we think AI has only one type of intelligence. But in reality, we can create diverse minds and intelligences.

By developing these various types of intelligence, we will come to understand that human thinking is not at the center, but rather on the periphery.
Future AI, as a distinct form of intelligence different from ours, will think differently from humans. This will create new innovations and economic spheres.

 

AI will spark a second industrial revolution

The second benefit of AI is that it will bring about a "second industrial revolution" in our society. The "first industrial revolution" was the advent of "artificial power" about 150 years ago.
In agrarian societies, production relied on human and animal muscle power or harnessing natural forces. Human and animal strength was limited, and production systems were constrained, requiring extremely slow processes to create large-scale items.

So we created "artificial power" using steam, electricity, and fossil fuels. As a result, we could power a car with 250 horsepower at the flick of a switch. We built skyscrapers, constructed roads, railways, and factories, and created refrigerators and televisions. Each of these developments multiplied millions of times over, sparking the Industrial Revolution and transforming our lives.

The first Industrial Revolution brought electricity to every home, residence, and farm. There was no longer a need to produce electricity oneself; it could be obtained from an outlet. In other words, the commoditization of electricity became the source of innovation.

For example, if you were a farmer, you could add electricity to your existing hand pump to create an "electric pump." By adding artificial power to any manual device, you could create an automated one.

The Second Industrial Revolution further evolves this electric pump by adding AI. This creates a smart pump. Earlier, I mentioned a car having 250 horsepower. This time, by adding AI, we might describe it as having "250 brains." The sum of this 250 intelligence and horsepower is the "self-driving car." In other words, it's the combination of the First and Second Industrial Revolutions.

The formula representing the business model of new startups emerging in the future will be simple. It will be the formula "X + AI," where you select something X and add AI to it.

Consider what happens when you add AI to shoes or chairs. Or what happens when you add AI to taxis? It might become Uber. Repeating this across various fields demonstrates just how significant the impact can be.

The future will see AI and humans working together

Robots are devices with AI embedded in their bodies. When older robots were used in factories, cages were built around them. This was because they were too clumsy and lacked sensors, posing a risk of injuring or even killing people since they couldn't detect them. However, smart industrial robots equipped with AI can detect humans and take care not to harm them.

More importantly, smart robots possess the ability to learn tasks by demonstration—simply show them "this is the job I want done," and they will imitate it. Furthermore, they autonomously learn until they can execute it correctly.

These smart robots can collaborate with humans. This is crucial. It will change how we define work. While many of our tasks will disappear, robots will create new jobs.

The jobs robots take from us are those where efficiency is paramount. Efficiency is a robot's forte, while humans neither excel at nor particularly enjoy it. Conversely, the work we excel at involves tasks where efficiency isn't critical—precisely the kind of work tied to innovation.
This is because innovation is inherently inefficient. Without numerous failures, innovation cannot happen. Art, creativity, and human relationships are also inefficient.

Garry Kasparov, the world chess champion, lost a match to a supercomputer equipped with AI. However, he realized that if he had access to all historical chess moves in real time, just like the AI, he could have won. So he launched a new league where AI and humans play chess together.

In this league, humans can play independently against AI or collaborate with AI. He called these teams of humans and AI "Centaurs." Today, the world's best chess player is neither AI nor human. It is the Centaur: a human plus AI.

This Centaur will become the model for the future. The best doctor is a human physician plus AI. That means humans working side-by-side with robots. Human and robotic intelligence are complementary; collaborating with robots yields better results than either working alone.
This is the "COGNIFYING" world.

※Continued in Part 2

You can also read the interview here on AdTie!

 

Additionally, a special report on this event is available on cotas.
Planning & Production: Dentsu Inc. Event & Space Design Bureau, Aki Kanahara