Chapter 511: The Technology That Should Not Exist

Late September, Odaiba, Saionji Tower.

This black tower, repeatedly promoted by the Tokyo Metropolitan Government as the "new landmark of the Waterfront Sub-center," had been standing on the shore of Tokyo Bay for quite some time.

From the unfinished construction area of the Rainbow Bridge, the Shinagawa area, and even the opposite shore of Tokyo Bay on a clear day, one could see the gradually taking-shape high-rise building from afar.

However, "going into operation" in the true sense would still have to wait a while longer.

Although the main building was basically completed and most of the glass curtain walls had been installed, the internal office areas, public spaces, and surrounding facilities were still undergoing the final stage of construction.

Every day, a large number of engineering vehicles still entered and exited through the dedicated passage, and occasionally the sounds of metal cutting and equipment installation could be heard from the floors that had not yet been sealed off.

The places that had truly begun to go into operation were instead hidden underground.

Leading to the underground facilities was a dedicated elevator independent of the above-ground office areas.

After Satsuki and her party entered, the elevator went all the way down without making any stops midway.

It wasn’t until the fourth underground floor that the elevator car slowly decelerated.

Accompanied by a soft click, the elevator stopped.

Yet the doors did not open.

Starting from this floor, the areas further down were collectively referred to within Saionji as "Zone Zero."

The Tamatebako computation matrix center, the comprehensive experimental zone, and several technical facilities with higher security classifications were all located within it.

Being able to reach the fourth underground floor did not mean having the qualification to continue downwards.

Anyone entering Zone Zero had to complete an independent identity authentication here once again.

The originally seamless metal wall panel on the right side of the elevator recessed inward slightly, and then slid open silently.

A set of hidden biometric recognition devices emerged from behind the wall panel.

At the very top was a black observation hole only the size of a thumb, and below it was a slightly recessed glass panel.

Satsuki had obviously used it many times before.

She stepped forward and aligned her right eye with the observation hole.

A dim red shimmer quickly lit up.

The retinal scan lasted less than two seconds.

At the same time, Satsuki placed her right hand on the glass panel below.

The sensors hidden beneath the glass then began to read the palm shape, fingerprints, and several physiological data used for liveness confirmation.

A few seconds later, the red light went out.

A calm electronic female voice sounded inside the elevator.

"Identity confirmed."

"Saionji Satsuki."

"Privilege level: Zero."

The floor display screen, which previously had no response, immediately lit up.

Several groups of floor numbers that previously did not exist at all on ordinary elevator panels appeared in sequence.

Subsequently, the elevator restarted.

A slight feeling of weightlessness came from beneath her feet once again.

Satsuki stood back in her original position, watching the numbers on the display continue to jump downwards.

Negative five... negative six... negative seven.

Until the eighth basement floor, the elevator finally stopped slowly.

As soon as the doors opened, a completely different air from the outside rushed over.

The temperature was much lower than above ground, and even the sounds seemed to have been filtered by thick walls, leaving only the continuous low hum of distant ventilation equipment.

Chizuru stepped out a step ahead.

Satsuki followed behind and looked up at the top of the corridor.

Silver-gray metal wire troughs ran forward along the ceiling, and several pipelines marked with different colors connected to different areas respectively.

The left passage led to the already operational "Tamatebako" computation matrix center, and the right was an entire comprehensive experimental zone that had just completed the first stage of construction.

This place was a bit different from other research institutes of the Saionji Group.

It neither clearly corresponded to a specific subsidiary nor had production tasks.

Even when it could generate profits, nobody could say for sure.

Projects that could enter here only had two judgment criteria.

One was that Satsuki felt it was worth doing.

The second was projects deemed by the recently established Saionji Technical Committee as having "unclear commercial potential for now, but potentially huge impact once successful."

High risk, small quantity, while the budget and equipment conditions far exceeded the upper limit that ordinary corporate research institutes could obtain.

The things researched by many projects even looked absurd right now.

High-density energy storage materials, automated mechanical control, special display devices, and several laboratories whose specific research topics were unknown even to many top executives of the group.

Brain-computer interface was one of them.

...

Just as Satsuki turned the corridor, through the thick observation glass ahead, she already saw quite a few people standing inside.

In the center of the laboratory stood several instruments whose names she couldn’t call out.

Against one wall were several full rows of electronic equipment, with connection cables extending from behind the equipment cabinets all the way to the central experimental bench.

On the other side, separated by a transparent baffle, a small animal experimental area was carved out separately.

Emi stood near the cubicle of the animal experimental area, holding a clipboard in her hand, and was talking to a young man wearing the same white lab coat.

"Did the monkey eat anything this morning?"

"Normal food intake. Fruit and formulated feed were given at seven-forty, and everything was eaten up by eight-twenty."

The other party flipped through the records in his hand.

"Body temperature thirty-six point eight, activity level normal, no abnormalities in the skin of the implantation area."

"What about the impedance?"

"In the last test last night, forty-seven out of sixty-four channels were still valid. Among them, thirty-nine met the signal-to-noise ratio standard, which is two less than the day before yesterday."

Emi lowered her head and wrote a few strokes on the clipboard.

"The twenty-third and forty-first channels?"

"Right, those two. They started degrading last week and officially dropped below the threshold yesterday."

Emi nodded without further questioning.

She looked up at the rhesus macaque in the cubicle.

Through the observation window, one could see that the brownish-gray monkey was squatting on the platform of the enclosure, grabbing a metal rod fixed to the side wall with one hand, and slowly peeling off a small piece of banana peel with the other.

The hair in the implantation area on top of its head had been shaved off, revealing a grayish-white area the size of a coin.

The surface was covered with a thin layer of biogel, and the base of the connector protruded slightly from the middle, looking like a small button grown on the scalp.

The monkey didn’t care at all about the thing on top of its head.

It stuffed the banana into its mouth, chewed twice, turned its head to look at the observation window side, and lowered its head again to continue eating.

"Condition is good."

Emi clipped her pen back onto the clipboard and turned to walk towards the main operation area.

Several other researchers were already at their posts.

Someone was sitting in front of the signal processing terminal adjusting parameters, someone was checking the cables connected to the acquisition equipment, and a woman wearing anti-static gloves was carefully aiming a pair of tweezers at the positioning slot of a set of microelectrode probes.

"Mr. Kawakami."

Emi walked over to the signal processing terminal and called out to the man sitting there.

Kawakami Mitsuo raised his head.

Thirty-seven years old, he wore thin-frame glasses, and on the table in front of him were several recording papers covered with calculation formulas and neuron discharge data.

He originally studied motor control and neural networks at Osaka University. After transferring to ATR a few years ago, he was mainly responsible for research in human motor control, neural network learning, and computational models.

Last year, when Saionji began expanding the brain-computer interface project, Emi noticed his name while organizing related papers.

The two sides were initially just cooperating.

Saionji provided computing resources and research funding, while Kawakami was responsible for helping the team build motor control and neural decoding models.

As the project continued to expand, the time he spent in Odaiba became longer and longer, and now he had become one of the most important heads of the entire decoding group.

"Have the decoding programs run?"

"Run three times."

Kawakami turned the screen around to let her see.

"Both yesterday’s model and the version recalibrated this morning have been tested. The offline accuracy is stable at around seventy-eight percent. Online will be a bit lower, probably seventy to seventy-three."

"What about the latency?"

"One hundred and forty milliseconds."

Kawakami pushed up his glasses.

"Sixty milliseconds faster than last month’s version. Mainly because Tamatebako reassigned a group of computing nodes to us yesterday, so matrix calculations no longer have to queue with several other projects."

Emi nodded.

One hundred and forty milliseconds.

Three months ago, this number was still above two hundred.

A year before that, they didn’t even have a set of online decoding programs that could run stably for a long time.

Nowadays, what truly restricted the experiment had gradually shifted from computing equipment and software to electrodes, tissue responses, and those even more troublesome biological problems.

Emi glanced at the test results on the screen again, and just as she lowered her head to continue speaking, she saw that the safety door outside the glass had opened.

The voice stopped abruptly.

Others in the laboratory also looked over along her gaze.

Satsuki was walking in from outside.

The originally serious expression on Emi’s face visibly brightened for a moment.

But it was only for a moment.

There were still people from the entire project team around, and she quickly restrained her expression again, held her clipboard steady, and walked towards the entrance.

"Eldest Miss."

Looking at Emi, who was so rarely serious, Satsuki still felt it was somewhat novel.

"Are you ready?"

"Already ready."

Emi nodded seriously.

Compared to how the two of them usually got along privately, she looked exceptionally well-behaved today.

At least she didn’t pounce over while shouting Satsuki-chan.

Even her tone of voice intentionally maintained what a laboratory head should have.

Satsuki looked at her, and the corners of her mouth couldn’t help twitching slightly.

Quite convincing.

But with a dozen researchers standing around, she wasn’t prepared to undermine Emi at a time like this.

"Then let’s begin."

"Yes."

Emi turned around.

"This way."

She led Satsuki to the observation area.

...

After the blackout curtain of the cubicle was drawn open, Satsuki saw that rhesus macaque up close for the first time.

It had been transferred to a fixation device connected to the experimental equipment.

Its body was gently secured in the seat by flexible straps, with a limited range of motion, but its head and arms remained free.

Plugged into the connector base on top of its head was a slender signal cable, extending along the fixed arm all the way to the acquisition equipment outside the cubicle.

The monkey’s eyes were staring at a vertically installed LCD screen in front of it.

The screen had a black background with nothing on it.

"This is our experimental subject with the longest long-term implantation time so far."

Emi stood on the outside of the cubicle, her laser pointer pointing to the connector on top of the monkey’s head.

"ID MK-07. Motor cortex implantation, sixty-four-channel flexible polyimide microelectrode array, has been continuously working for fifty-three days."

Satsuki glanced at the channel status map on the display terminal next to it.

Sixty-four small squares were arranged in a matrix, most of which glowed green, a dozen were yellow, and a few had turned gray.

"Currently forty-seven active channels."

Emi also looked at that screen.

"Thirty-nine meet the signal-to-noise ratio standard."

"Can it still maintain this ratio after fifty-three days?" Satsuki asked.

"Yes, this is better than we expected."

Emi nodded.

"The other two in the same batch had already dropped below thirty around the fortieth day. MK-07’s individual differences might be more favorable, but it’s also related to the surface coating improvements of this version of probes."

Satsuki did not continue to ask; her attention had returned to the screen in front of the monkey.

Emi turned to Kawakami.

"Start."

Kawakami clicked the keyboard.

A small white dot appeared right in the center of the black screen.

Immediately afterwards, a yellow square marker lit up in the upper right corner of the screen.

The monkey’s line of sight was immediately attracted over.

It stared at that yellow marker for less than a second before its gaze returned to the white dot in the center.

Then the white dot moved.

Very slowly, somewhat hesitatingly.

It first moved a short distance to the upper right, paused for a moment, and then shifted a bit to the left, seeming to hesitate on the direction.

Then it re-corrected and began moving towards the yellow marker in an unstable but generally correct trajectory.

A few seconds later, the white dot touched the edge of the yellow square.

The screen flashed. The yellow square disappeared, and reappeared in the lower left corner with a new position.

The white dot began to move again.

Satsuki stood outside the cubicle, watching all this through the glass.

The monkey’s hands always rested on the armrests of the seat, touching nothing.

Its fingers occasionally clenched, like some residual grasping reflex.

But what was truly driving that cursor on the screen were the neural discharge signals collected from its motor cortex—through sixty-four probes thinner than a strand of hair.

The signals were transmitted via cables to the acquisition front-end outside the cubicle, processed by amplifiers, and then sent into the decoding terminal.

Kawakami’s algorithm translated the activity patterns of dozens of neurons into coordinate instructions within one hundred and forty milliseconds, and then pushed the white dot on the screen.

The entire process involved no mechanical devices participating.

The cursor touched the yellow square in the lower left corner. The screen refreshed again, and a new marker appeared directly above.

The monkey glanced at it, and the cursor followed upwards.

This time it moved a little more smoothly.

"During continuous testing, about seven out of ten times it can hit the target within the specified time."

Emi’s voice came from the side.

"Complex trajectories will be a bit worse, especially paths requiring sharp turns; the false-positive rate is still relatively high."

Satsuki did not turn back. She was looking at the crooked, stop-and-go white dot on the screen.

"The mechanical arm is also connected?"

"At the console next door."

Emi gestured toward the right.

"The same decoding output was split into two paths. One controlled the cursor, and the other simultaneously drove a simple 3-DOF mechanical arm."

Satsuki walked over and took a look.

A silver-gray small mechanical arm was fixed on the clamp of the console.

Its front end was equipped with a simple gripper, which was currently making corresponding movements following the neural activity of the monkey in the cubicle.

Arm stretches to the right — cursor moves to the right.

Arm probes forward — cursor moves upward.

There was a visible synchronous relationship between the action and the cursor, but the movement of the mechanical arm was stiffer than the cursor.

Every time the direction switched, it would pause first, as if it needed to reconfirm the instruction before daring to continue.

"Reaching forward and left-right are okay."

The researcher in charge of the mechanical arm part chimed in from the side.

She took off her anti-static gloves and turned a page of data.

"The accuracy of the retraction action is relatively poor, only about fifty percent or so. The opening and closing of the gripper can currently only do two gears: fully open and fully closed, and the middle strength cannot be controlled."

Satsuki stood in front of the mechanical arm and watched for a while.

The gripper opened again, stopped in the air, and then slowly closed. The movement was not smooth, and a very light gear sound was emitted when the joints turned.

This time, the smile on her face gradually faded a bit.

Not because she was dissatisfied.

Quite the contrary.

Although she was far from a professional in brain science, when she was doing finance in her past life, she had read countless industrial reports, laboratory intelligence, and paper abstracts in order to judge the value of tech companies.

Brain-computer interfaces, things that were repeatedly hyped by the capital market later on, were naturally something she had come into contact with as well.

Which specific paper was in which year, she couldn’t remember clearly anymore.

But she still had impressions of some key nodes.

Monkeys directly controlling external devices relying on motor cortex signals.

Multi-channel invasive interfaces forming a real-time closed-loop in the true sense.

These things...

Absolutely should not have reached the degree before her eyes back in 1991.