Chapter 1813

Compared with normal users, the number of devices that Gu Qing wears is indeed very large, just from the perspective of pure quantity.

For example, the improved SR1 glasses that he wears for a long time, the tiny devices in the ears and nasal cavity, the health bracelet on the wrist, and some safety instruments and equipment...

Conservative estimates are all above double digits.

As for the personal smart terminal, this product that is about to be released commercially, although he does not have the highest authority version of the neural connection of the personal smart terminal, he is also a personal user of the standard version.

After all, the highest authority version of the neural connection of the personal smart terminal will only be introduced at the press conference at the subsequent press conference, and will not be sold.

This version of the personal smart terminal will be the exclusive equipment of the mecha driver, allowing the elites who operate and control the mecha to better control the mecha with neural connection technology.

Soon, Xuanwu sent all the experimental data of the personal smart terminal to Gu Qing, and also compiled a copy of the content of the personal smart terminal project mentioned in the logs of the experimenters of other related projects and provided it to Gu Qing.

As for Gu Qing's statement that "if this product really has technical risks, we may need to optimize it again." Xuanwu did not make any statement or feedback.

But it doesn't know that it is precisely because it has no feedback that Gu Qing thinks this is the most important feedback.

Nodding and shaking your head are both good choices.

But those who are very cunning often respond to all changes with silence.

[Xia Yin Star Calendar September 11, 1024, Yanjing time 9:36 am... A12 experimenter suffered from neurasthenia...]

[Xia Yin Star Calendar September 11, 1024, British time 4:25 pm... C71 experimenter was diagnosed with epilepsy and facial nerve disorder...]

[Xia Yin Star Calendar January 22, 1025, Yanjing time 3:33 pm... A192 experimenter began to suffer from neurasthenia after ten days of insomnia and dreaminess...]

[Xia Yin Star Calendar January 26, 1025... A192 experimenter suffered a heart attack after ten days of insomnia and dreaminess and two days of neurasthenia...]

From the logs and data of the laboratory in Daxia to the logs and data of laboratories in other regions abroad, Gu Qing's eyes remained serious.

"Almost all participants in the experiment will have a transition period of neurasthenia before the process materials are improved. The current material progress, that is, the harmless feedback in the past six months, can only be adapted by the native Chinese people of Daxia.

It's not just adaptation, it should be said that it can reach a certain coordination with the chip materials of personal smart terminals in the shortest time.

In the research of the life research department, the genetic source of the native Chinese people of Daxia can be traced back to the three most primitive ancestors. After so many years, it is still shining..."

Looking at the experimental log in front of him, Gu Qing's thoughts became more and more chaotic.

There are cases where the experimental data has become excellent due to material upgrades; there are cases where there are obvious changes due to the different physical conditions, nationalities, and blood of the experimenters; and there are also cases where the personal smart terminal is installed, and after the rejection period, some overseas death row prisoners who suddenly suffered heart attacks and strokes...

"In terms of technology alone, personal smart terminals equipped with support for smart program AI collaboration have the highest mortality rate without exception in the early stages of the experiment.

Even if the materials of chips, sensor devices, and neural connectors have made epoch-making progress, as long as the smart program AI is loaded into the personal smart terminal, subsequent tests will still affect the user.

And the personal smart terminals that Daxia officials can agree to our commercial sales not only do not have access permissions for smart program AI, but also cannot have any complex programs loaded.

Only simple operations are allowed to connect smart home appliances of all smart systems in Jiuzhou, allowing users to wear SR1 or K1 helmets without hands, and use neural consciousness to simply operate, see virtual content, see through the external environment, and realize the integration of augmented reality and the real world..."

With emotion, Gu Qing was not very satisfied with this first generation of commercially available personal smart terminals.

At first, he planned to make this product even better, and then sell it directly to these wealthy elite users at a high price.

Now it seems that the Daxia official has many concerns and worries about neural connection technology, and at the same time does not want Jiuzhou Technology's products to be "out of touch with the masses".

However, Gu Qing understands this decision.

But he also thought of making money in other ways for the loss in this aspect.

After his thoughts diverged to this point, Gu Qing did not continue to diverge.

He immediately retracted his thoughts and began to seriously study AI.

And this time, he did not do a review, but temporarily ignored his own knowledge system technology, and began to study Xiawei Technology's Pangu Big Model and OpenAI's ChatGPT in his personal database.

"The founder of Xiawei Pangu Big Model is a scientist in the field of cloud artificial intelligence. After choosing the two tracks of NLP and CV at the beginning, he established three core design principles.

First, the model should be large and can absorb massive amounts of data, so that there can be enough experience; second, the network structure should be strong to truly bring out the performance of the model; third, it should have excellent generalization ability and can be truly implemented in work scenarios in all walks of life.

These choices are rational. Xia Wei has been deeply involved in the field of AI for many years, has accumulated a lot of full-stack technology, and has sufficient resources.

From forming a team to having a core team of experts, project establishment, and developing a very large-scale pre-trained model in one year, it is indeed not easy.

Wait, Xia Weiyi started with a visual pre-training model with 3 billion parameters, and a Chinese language pre-training model with 100 billion parameters and 40TB training data jointly developed with Circular Intelligence and Pengcheng Laboratory?

With more than 20 PhDs, more than 30 engineers, and dozens of other experts, this research and development technology is relatively fast.

On September 23, Xia Yin 1021, Xia Wei released a large model of Xia Weiyun's Pangu drug molecules. I actually have no memory of this?

Xia Weiyun and the Pudong Institute of Materia Medica of the Daxia Academy of Sciences jointly trained the Xia Weiyun Pangu drug molecule large model to pre-train the chemical structures of 1.7 billion drug molecules that actually exist on the market. Under the chemical unsupervised learning mode, The achievement of structural reconstruction rate, legality, uniqueness and other indicators are all better than existing methods...

At that time, the Xuanwu I developed was still analyzing related materials in the field of hardware semiconductors? "

Gu Qing suddenly discovered that an AI program responsible for sorting out daily news for him seemed to have some problems with its work.