Crush - episode 6

(Continued from episode 5)

Later, people called that day Red Saturday.

Newell was extremely adept at execution, full of confidence in his technical skills, to the point of arrogance. He acted arbitrarily within the company, and no subordinate dared point out his mistakes. While this allowed him to efficiently advance projects, it left a huge hidden danger in reliability.

The night I sent the message, a major accident occurred at an air valve in Mars City No. 2, where Xiao-Hui lived, leading to a large-scale leak. The monitoring system quickly detected the problem but failed to issue a timely warning due to Newell's extreme distrust of AI - he insisted that red alerts must be manually confirmed. The on-call engineer, fatigued from weeks of continuous work due to the last round of lay-offs, slept through the alert. When he woke up struggling to breathe, it was too late.

The pressure dropped dramatically from near one standard atmosphere to less than 0.2 in just five minutes and further dropped to 0.05 ten minutes later. With the dissipation of oxygen, over a thousand residents, except for a few Skyway employees wearing spacesuits outside the city, all suffocated to death.

When the news reached Earth, Skyway’s stock plummeted by 99.3%. Several despairing shareholders leapt from high buildings, adding to the death toll.

My AI assistant pushed this breaking news to me at three in the morning. I stared at my phone in disbelief, as if struck in the chest. That damn Lao San's a jinx!

When I got over the shock, I took out a bottle of Tuopai liquor from the cabinet and went to the rooftop.

In the night breeze, I looked up for that orange-red star. It was still there, unmoving, as if nothing had happened. Xiao-Hui, why did you leave? If I had been admitted to Renmin University the same year as you, would you have been with Xu the Poet? Would you have gone to Mars?

I moved my lips but couldn't make a sound. I put my hands together and silently gazed at Mars for a long time. Then I poured the bottle of liquor onto the ground.

Xiao-Hui, the Tuojiang River will never dry up.

Cloud

The mega accident on "Red Saturday" caused the death of over a thousand people in Mars City No. 2, thoroughly extinguishing the public's enthusiasm for Newell's Martian colonization project. People had realized the fragility of the flesh: life based on carbon compounds has strict requirements on the environment. Planets suitable for carbon-based life are beyond rare in the universe. Humanity's hope of finding a second home in the universe may be wiped out thousands of times before AI even has a chance to strike.

The study of digital life has been brought to the agenda: If humans could escape the constraints of carbon-based bodies and exist in digital form, wouldn't that greatly increase the chance of survival in the universe?

After graduating from university, I had been working on research and development at the Institute of Computing of the Chinese Academy of Sciences. Around 2023, artificial intelligence suddenly became a big hit, and I began to study artificial neural networks.

By 2040, when Xiao-Hui went to Mars, the connections between neurons and their strength could be precisely measured by CT. Theoretically, a model of a person's brain could be built from the scans. If you run the model and feed it the neural signals of the sensory organs, it can output the person's reactions.

Would such a simulated human have its own consciousness? This question is both exciting and frightening. If it has consciousness, should it be given human rights? Is it inhumane to interrupt the model’s operation?

Until legislation caught up, these issues were off limits to research. The consensus among most AI experts was: Humans cannot play God. A wrong step in such a grave matter may have unimaginable consequences, and utmost caution must be exercised.

～～～～

After Xiao-Hui's tragic death, on every anniversary I would log on to her virtual cemetery and place a bunch of her favorite gardenia flowers. In my childhood, during hot summer days, old ladies would wander the streets selling gardenia flower bracelets, one string for a dime. I would never forget the white flowers on Xiao-Hui's arms and their lingering fragrance.

At our middle school graduation ceremony, Xiao-Hui sang Tsai Chin's "Just Like Your Tenderness," wearing a gardenia necklace. Her ethereal voice struck my heart with every word:

Year by year passes by, I cannot stop reminiscing. Reminiscing you, reminiscing the past. Wish the sea breeze rises again, only for the hand of the waves, just like your tenderness.

After the performance, I found Tang, who accompanied Xiao-Hui on the guitar: I want to learn the guitar, can you teach me?

Everything had faded away. Year after year, I left messages at Xiao-Hui's virtual cemetery, but I knew she couldn't see them. Those questions that had plagued me for years might never be answered.

In 2061, I was already a 90-year-old man, ready to retire.

That year, I was leading a team at the Institute of Computing to collaborate with the Institute of Animals on a project called Zhihua ("intelligentization,") intending to digitize human consciousness. We had all the equipment, but due to policy and regulation, we were unable to conduct human experiments and had only tested on a few chimpanzees.

Our idea was simple: to use Tensor Processing Units (TPUs) with large computational power to run a neural network model of the chimpanzee brain, predicting its reaction to various stimuli. For ease of observation, we connected indicator lights to the virtual chimpanzee's brain neurons responsible for different functions. When the excitement of a neuron exceeded a threshold, the corresponding light would turn on, creating a ritualistic feel.

The project's name, Zhihua, was my proposal, in homage to a singer-songwriter I used to listen to when I was young. His name was Zhihua Zheng, and he had a widely-known song called The Stars Light Up, which sounds the same as The Chimpanzees Light Up in Mandarin.

After several adjustments, we achieved an 87% similarity between the virtual and real chimpanzee reactions. Considering the inherent uncertainty of biological systems, this high similarity far exceeded our expectations and could be deemed a great success.

However, the first generation of virtual chimpanzees still had significant differences from the real ones: their brains were rigid, parameters never changing; they could not form new memories, change existing habits, or learn new skills. Over time, the behavior of digital chimpanzees diverged more and more from the original.

Worse, the digital chimpanzees lacked the real chimpanzees’ passions and desires, and more resembled a walking corpse. Over time it even began to exhibit symptoms of depression.

Professor Qian, the Institute of Animals' mating expert, suggested that the electrical conductivity between neurons not only depend on the structure and spatial relationship of the neurons but also on the concentration of various hormones and electrolytes in the brain. For instance, endorphins cause physiological comfort, dopamine leads to psychological pleasure, all of which affect brain function. Our failure to provide the virtual chimpanzees with endorphins and dopamine disrupted the harmonious unity of the system, an utterly inhumane approach.

Qian pointed out two ways to us:

First, gradually enhance the strength of connections between frequently used neurons, and weaken the connections that are not often used. This simulates the animal's learning process.

Second, calculate the secretion of hormones based on brain responses, and then change the strength of neuron connections from constants to functions of various chemical concentrations in the brain.

Qian was indeed a professional in mating matters. With these improvements, the simulated chimpanzees’ intelligence and emotional intelligence caught up with the real chimpanzee. They could also maintain their youthful brain capacity and learning ability, not affected by dementia.

If this technology were applied to the human brain, it could not only overcome the frailty of carbon-based life but also achieve immortality. This would be humanity's singularity.

Everything was ready, except for the right opportunity. This opportunity was the state’s approval for human experiments.

Once human experiments succeeded, I could achieve digital immortality, having infinite time to find the answers to my questions.

But could I live to that day?

(Continue to episode 7)