After thanking the waiter who brought the cold drink, Atletico turned to his wife.

"There won't be any ocean empires, humanity didn't give them that time. ”

Due to hundreds of years of commercial whaling, as well as environmental pollution and the destruction of food resources by human fishing, the number of various cetaceans is declining dramatically. As a result, from 1986 onwards, countries gradually reduced or even banned commercial whaling, leaving neon in Asia and Denmark in Europe out of the picture.

In the name of scientific research, fishermen kill more than 2,000 cetaceans a year, dissect them, weigh them and map them, and then sell them on the market, including both whales and dolphins (cetacean dolphin family). The locals of the Faroe Islands, the latter territory, also adhere to the tradition of killing dolphins and selling them.

Just as sharks are in decline because shark fins are in decline and pangolins are endangered because of their scales, cetaceans are also victims of "buy and sell", and as long as humans need their skin, muscle, fat and bones, these animals will die.

Tang Jiayi sighed: "Yes, if you continue to hunt and kill indiscriminately, dolphins and whales will be wiped out, and naturally there will be no chance to evolve intelligence that is enough to defeat humans, and the Ocean Empire will be even more shadowless."

"Even if humans don't kill them, it's hard for them to live freely," Atletico shook his head, "Food shortages, boat disturbances, marine pollution, and abnormal weather are also threatening them." Evolution can't keep up with the rapid change, and it's futile to help. ”

"Yes!" Touched by the sentence "Evolution can't catch up with change", Tang Jiayi's mood suddenly fell, "In fact, not only dolphins don't have much time, but there is also not much time left for humans, and the 21st century is likely to be the last human century." ”

Atletico smiled, "Time is really running out, there are only two minutes and thirty seconds left!"

On January 20 this year, Trump was officially sworn in as the 45th president of America. As a "congratulator" on his appointment, the country's Bulletin of Atomic Scientists set the famous "Doomsday Clock" a week later, indicating that humanity was thirty seconds closer to destruction.

After the end of World War II, a group of nuclear physicists involved in the construction of the atomic bomb created this organization and erected this special clock in 1947. Today, the watch is run by a group of Nobel Prize-winning atomists and environmentalists. Trump's immediate withdrawal from the Paris Agreement on climate change, as well as the renewal of his nuclear arsenal and the expansion of his intelligence agencies, clearly made these top scientists see the danger and react violently.

For the past 60 years, the hands of the Doomsday Clock have oscillated in an attempt to alert the public to international events that would prevent humanity from being destroyed by the madness of a few. In 1953, the United States and the Soviet Union competed to conduct high-power hydrogen bomb experiments, and the doomsday clock was set to two minutes, causing a huge sensation. Later, as the nuclear powers gradually canceled live-fire tests (using supercomputing simulations instead), the countdown was moved to more than three minutes, and now it is back to the "two-point era".

Of course, whether it's two minutes or three minutes, it's just a metaphor for stimulating the eyeballs. If someone really pressed the nuclear button, it would take about half an hour for the mushroom cloud to appear on Earth - this is the flight time of a ballistic missile, and it cannot be further accelerated, otherwise it will exceed the first cosmic velocity, and then fly out of the Earth's orbit and become space fireworks.

Of course, if that happens, no one will rejoice in the extra half hour.

Tang Jiayi was annoyed, her eyes widened, and she said in an urgent tone: "That's not what I'm talking about! Before the balance of terror is broken, the probability of a nuclear crisis breaking out is actually very, very low, but the 'AI crisis' is just around the corner." ”

"Is it close?" asked Atletico disapprovingly.

"Of course! It took 3 million to 5 million years for Australopithecus to evolve into Homo sapiens, and another 200,000 years for Homo sapiens to evolve into modern humans. Human evolution is measured in 10,000 years, and in the last few thousand years, there have only been changes in appendix, wisdom teeth, and height, and everything else has been completely unchanged. Sixty years ago, a computer the size of a house could only perform 1,000 calculations per second, but now chips the size of a fingernail can perform 3 billion calculations per second, and Go AI beats the top human players. What about 60 years from now? With today's development trends, it is only a matter of time before technology explodes and super AI emerges. Stephen Hawking, Elon Musk, and Ray Cozwell have all seen it, don't you feel it? Not nervous at all?"

"Hey!" Atletico laughed again, "I'm not nervous. Being the last generation of Homo sapiens is actually not a big deal. If the 'singularity' in Thunder's mouth really comes, it would be nice to download the memory and consciousness into the machine's body at that time, but it would be just a change of skin and body, and there would be no illness or disaster and no death, wouldn't it be beautiful?"

"By the way," Atletico looked at each other with a smirk, "don't you think it's strange that you put Hawking, Musk, and Coatswell together as arguments? And then there's Musk, who is now starting to support the theory of integration. ”

"Uh!" Tang Jiayi blushed, so she had to turn her grief and anger into food, and scooped a piece of ice cream into her mouth, although her teeth trembled and the corners of her mouth trembled, she still swallowed it.

Theoretical physicist Steven Haking, high-tech entrepreneur Elon Musk, and futurist Ray Kurzeil share the same view that AI is bound to rise. But when it comes to how to treat AI, the attitudes of the three are very different.

Hawking is very wary of the rise of AI, and he is worried that AI will take over the world in the future, or even end it.

If there really is a purpose-driven, absolutely rational artificial intelligence in the world, it will not know how to compromise, and it will not be bound by morality and law, and it is likely to make some crazy moves in order to complete the task efficiently, such as hospital management AI may kill all disabled people, patients, and genetic disease gene carriers in order to achieve the goal of minimizing medical expenses, and then make up for the gap with "perfect humans" made by bioengineering.

Musk's views are somewhat similar to his, and the two have called for a ban on the development of military intelligent robots for fear that these devices will eventually rebel against humanity. But as an entrepreneur or businessman, he is not very resistant to the arrival of artificial intelligence, with a large number of industrial robots installed in his gigafactories, and Tesla's Autopilot technology has been touted as a selling point — until a series of collisions have led to a downturn and the name of Autopilot.

As for Cozwill, he's more of a best-selling author, the kind of guy who doesn't talk to death.

He transplanted Moore's Law, which doubled the performance in the chip field in 18 months, to the software field, and then expanded to the biological field, and put forward the concept of "accelerated return" and "2050 singularity", believing that technological progress will bring about the exponential and rapid development of human society, and then become unrecognizable around 2050, and there will be strong artificial intelligence and super-artificial intelligence that surpass humans. At that time, there are only two paths in front of humans, either merge with the machine in exchange for eternal life, or be wiped out by the machine for various reasons.

Cozweil's theories have a lot of market among the general public, whether they are optimists who expect to embrace AI for eternity, or pessimists who are bearish on the future and fear AI, they are convinced of his "law of accelerated returns", which is historically proven, isn't it?

However, Atletico is well aware that human civilization and AI technology are indeed accelerating, but not as fast as he imagined, and his theory is too idealistic. He himself is the best counterexample, but unfortunately he can't publicly refute it.

Speaking of which, Atletico Madrid is barely the first "fusion man" who fuses humans with AI and machines, and the artificial intelligence E7U from outer space has transformed his body into a biological machine, with powerful storage and computing capabilities, and the indicators of strength, speed, and endurance have broken through the pinnacle of human beings. But that's all,Limited by hardware conditions,E7U's transformation of the body is not thorough,At the same time, its intelligent performance is also very average,It has long been tried by Atletico to find a way to be hollowed out,Reduced to a cracking research object。

The same applies to other artificial intelligences, whose dependence on hardware and energy makes them not as easy to come as optimists think, nor as powerful and invincible as pessimists think.

Based on Moore's Law, Cozwell is optimistic that the cost of computing is falling exponentially, and he proposes that when people can buy 100 million calculations per second for 1,000 dollars, strong artificial intelligence that surpasses humans will appear. Let's not talk about how to solve the algorithm problem, but only talk about the electricity bill. The peak computing power of Tianhe2 is 340 million times per second, and the power consumption is 20,000 kilowatts, which is the result of using low-temperature tap water provided by Huacheng Water Plant to cool down. Dear, 7060 kilowatts of billions of supercomputing, ordinary people can afford to use it?"

"What's wrong with this? The power consumption of computers is constantly decreasing!" Tang Jiayi pouted, "How big is the supercomputing volume and how high is the power consumption of 100 million times per second in the past? How big is the current mobile phone chip? The power is not even 5 watts!"

Atletico didn't bother to explain to her the difference between the use and function of supercomputing and embedded chips, and said: "The computing power per watt of chips is indeed improving, but there is a limit to this thing, otherwise the major fabs will not worry about how to solve the leakage problem and how to beautify the process figures." ”

"In short, Moore's Law is gradually becoming invalid in the semiconductor field, and it often hits a wall in other industries. Not to mention anything else, the energy field is dissatisfied with all kinds of slaps in the face. Thermal power generation has been around for 100 years, and the engine power is getting bigger and bigger, but the improvement of energy efficiency is very slow; solar panels have long been used in satellites, and the comprehensive production cost and power generation efficiency are still very poor; lithium-ion batteries have been out of the laboratory for more than 20 years, and the manufacturing cost has been significantly reduced, but the energy density has grown slowly, and it can be charged faster and used faster. ”

"There is also controlled nuclear fusion power generation, which has been known for decades, and now it can only be found in the laboratory for a very short time, and commercial power generation is still far away. Even if it does, there will be anti-nuclear fighters who want to drive the nuclear power plant far away, and even if fusion is safer than fission, they will still worry about the radioactivity of nuclear materials such as deuterium, tritium, and helium-3. This, combined with the cost of long-distance transmission, makes it almost impossible to bring down electricity bills. If the electricity bill cannot be lowered, the computing power of hundreds of billions of times will not be popularized, and the rise of AI will be impossible to talk about. ”

In other words, even if there is a super AI that is against humans, the leaders of various countries can make the super AI stun or even shut down as long as they send special forces to blast the cooling water pipes in the machine room or the field transmission tower. If it really doesn't work, you can press the nuclear button and eliminate large-scale power generation and substations at designated points, which will inevitably solve the problem. Because of the trouble of regulation and control, the power system is the sector with the highest degree of automation and the trend of networking, while strategic nuclear weapons that emphasize survivability and security and stability are bound to have the lowest degree of networking, and will become the last talisman in the hands of mankind at that time. Although it is a bit sad to collectively enter the age of the internal combustion engine or even the Stone Age, it is better than being enslaved by Skynet, right?"

Looking at Atletico Madrid, who was seriously deducing the man-machine battle, Tang Jiayi slowly showed a smile on her face and put down the spoon in her hand, "Since your view of the future is different from everyone else's, why don't you pass on your point of view?

"Ahem," Atletico raised his head and shook his head with a smile, "There's no need, no matter what outsiders say, our research won't stop, because it's really useful." Besides, this kind of atmosphere can also play a role in confusing opponents, so why rush to clear up?"

"You!" Tang Jiayi glanced at him with disgust, "At this stage, we should pay attention to social impact and be ready to lead positive energy at all times, how can we care about these small profits?"

"Let me be like someone, put on a pullover sweater, and say that I don't love money and haven't been paid for more than ten years?

Tang Jiayi once again entered the "Huafei mode", "You don't have to learn from him! It's not bad to learn from Gates, Musk, Zuckerberg, and Bezos!"

Atletico nodded, "Let me be like Gates and urine (actually multi-layer purified)?"

Noticing someone's expression on the verge of going berserk, he hurriedly ended the flirting mode, and said, "Let's not talk about that, let's talk about your new game and what type of game you plan to make." ”

"Of course, it's an idle mobile game with a cute art style!"

"A copy of 'Cat's Backyard', can you pursue it a little!"

"Cat Backyard" is a neon mobile game, players need to manage a yard, place food bowls, water basins, cat litter boxes and other facilities, and then there will be all kinds of cats constantly visiting. Those who like it find it tranquil and interesting, and those who don't like it find it simple and boring.

"We are very pursuing, in addition to more cute and beautiful cats, we also plan dogs, pandas, ornamental fish and ornamental birds, the content is much richer than the cat's backyard. ”

"Change the soup but not the medicine. ”

"And what do you think?"

(End of chapter)