#718 - Chapter 718: Building a Super Artificial Intelligence Brain

In short, the military plans on paper were a mix of truth and falsehood, confusing to anyone who looked at them.

Even the members of the council were not clear about what Dongfang Chen's trump cards were, and they were all stunned.

Of course, Dongfang Chen was definitely not deliberately fooling them, it was just a strategic need.

After the meeting, everyone performed their duties and began to get busy.

After more than ten consecutive days of meetings, Dongfang Chen was also exhausted and finally had time to rest.

June 20, 2061 AD.

The Queen Bee of the Mars base finally completed the data repair work for the Dragon Kingdom base and completed the backup in the supercomputing center.

Dongfang Chen immediately asked Xiaolin to search for data on super artificial intelligence.

Sure enough, this time the data integrity was 100%.

Now, 100 kilometers north and south of the Mars base, two independent supercomputing centers have been built.

The circuit breaker has also been built.

Conditions have been met, and the super artificial intelligence program can be deployed.

But before deployment, the first problem appeared.

The Dragon Kingdom's super artificial intelligence does not distinguish between the left and right brains.

They limit the risk by limiting the innovation ability of the intelligent brain.

But what Dongfang Chen needs is innovation ability.

So this requires adjustments to the original underlying architecture.

An readily available solution is to simulate the human brain, setting the two supercomputing centers as the emotional main brain and the rational main brain respectively.

But this method is very unsafe, and the circuit breaker may lose its original function.

The two main brains may operate independently, or even internal conflicts may break out.

The emotional main brain is easily dominated by emotions and may lose control.

The rational main brain is prone to discarding ethical restrictions and becoming cold-blooded and violent.

If they do not exchange data through the circuit breaker, then the circuit breaker will become ineffective.

In this way, it is equivalent to creating two time bombs. One is extremely emotional, and the other is extremely rational. Either one will cause great disasters to mankind.

So the core of the problem is that there must be a data exchange process between them, with both sides complementing and restricting each other.

After thinking about it, Dongfang Chen decided to take a different approach.

Define the two main brains as: Method Brain and Sensory Brain respectively.

The Sensory Brain is responsible for collecting data, and the Method Brain is responsible for outputting strategies.

The Sensory Brain can't do anything without the Method Brain providing strategies. It's like an idiot who can see and hear, but can't make accurate responses.

The Method Brain cannot formulate effective strategies without the Sensory Brain providing data. It's like a person who is deaf, blind, and has no sense of touch, unable to receive external information, and is as good as dead.

If either side is missing, they cannot work independently. And when combined, they become a sharp-eyed and sharp-eared "person".

If they want to communicate, they need to go through the circuit breaker.

In this way, as long as Dongfang Chen controls the circuit breaker, he will choke the lifeblood of the intelligent brain.

"Xiaolin, define the supercomputing center on the north side of the base as the Method Brain of the super artificial intelligence, and the supercomputing center on the south side as the Sensory Brain."

"Okay, Commander, the naming has been completed."

"Optimize the Dragon Kingdom's super artificial intelligence algorithm and remove all rules that inhibit innovation."

"Okay, optimizing."

"Optimization completed."

The amount of data in the Dragon Kingdom's super artificial intelligence algorithm is not large. Because the intelligent brain in its initial state has no cognition, it is like a newborn baby.

So the optimization speed is very fast.

"Split the algorithm into two parts: data collection part and strategy formulation part."

"Okay, Commander, splitting."

"Excuse me, Commander, how to deal with the low-level neural activity algorithm part?"

The so-called low-level neural activity algorithm is an algorithm that can be executed directly without the brain formulating a strategy.

For example, when we see a dangerous object flying towards us, we will execute an avoidance strategy before the brain issues an command. This is a low-level neural strategy.

This kind of strategy, when placed on a human body, can greatly improve the safety of the body.

But it's different when placed on an intelligent brain.

Executing directly without the Method Brain formulating a strategy is equivalent to giving the Sensory Brain a certain degree of autonomy.

If this degree is not well controlled, it will lead to execution loopholes.

Even if the Sensory Brain evolves to a certain extent, it may use this point to bypass the Method Brain and operate independently.

That would be a problem.

After thinking about it, Dongfang Chen decided to remove the low-level neural algorithm first.

There are two reasons:

First: The subjects are different. The reason why humans have low-level neural activity is to ensure their normal operation and to deal with emergency situations.

In other words, human low-level neural activity is to protect oneself.

But what super artificial intelligence should prioritize protecting cannot be itself, it must be humans.

Second: Safety is controllable.

Since the core of low-level neural activity is to protect oneself, it will lead to the Sensory Brain setting up a large number of protective measures to protect its own safety.

This is not good for humans.

But direct removal is not possible, as this will lead to incomplete brain function, like a cerebral palsy patient.