The Chinese military uses DeepSeek in non-combat missions.

The Chinese People's Liberation Army, the PLA uses DeepSeek AI in non-combat missions. The Deep Seek can operate as the opponent and data seek missions in war games or military training. 

The AI can be a good opponent for generals and admirals to research and develop tactics. That they want to use. 

To be effective generals and admirals those people need good opponents in their war games.  AI requires lots of information that they can benefit from in those missions. The DeepSeek can also used to train other AIs for military missions. 

The problem is that. Open military AIs can be banned.  And if mission-critical AI operates directly with outside users. 

Those users can infect it using malicious code. And if that AI does not work right. That is a catastrophe. 

And there is the possibility. In real cases, malware can infect those mission-intensive systems. Or as an example, electronic counter measures, ECM, and denial of service, DOS attacks can deny those LLM's operations. 

So, the solution is to use open applications as gloves. That keeps the military AI sterile against malicious code. 

If mission-critical AI operates directly with outsider users. They can see things like its server gates. The server that runs the LLM is similarly vulnerable to other servers. Hackers and malware can make attacks against those servers. 

Malicious software can infect that large language model, LLM. 

The open civilian application makes one extra layer between outside users and the military AI. 

The DeepSeek can collect databases from multiple sources. And when it is ready. 

The system can transfer those databases to other AI. The thing is that. The AI can use multiple sources to develop or train itself. 

The training is like collecting some kind of encyclopedia of things that the AI must react to. Without information the AI is helpless. It must recognize details and then compile a reaction database with things that it sees. 

The AI must recognize the situation and then find the action that allows it to make counter movement for that thing. 

Those data sources can be sensors like spy satellites and recon planes. The system can also use things like computer games to create counteraction for some maneuvers. The AI can transfer the maneuvers that it detects from those sensors to the computer game. And then, the opponent who might not know about that information source starts to play against those maneuvers. 

The system can render those objects like tanks to things. That even Western players can accept them. That kind of network-based data-collecting and training systems can teach those military AIs at a level. That they will not otherwise reach. The large-scale information collection is the thing that makes the AI more intelligent and capable to give response multiple situations. 

https://interestingengineering.com/military/china-pla-deepseek-ai-non-combat-operations

Image: Interesting Engineering

Researchers say that DeepSeek training costs were 1,3 billion, not 6 million.

When we talk about China and its authoritarian commanding system. We must realize. That the costs that authoritarian governments and companies that cooperate with them are what the government wants. In China, there is no business without government. The real costs of AI training might form in different ways in China than in Europe. Those ways might be different than we follow in Western countries. 

If governmental offices like intelligence and the People's Liberation Army participate in the AI projects. Those offices are not telling their salaries and costs. There is also suspicion that the Chinese military intelligence offered data, that it gets from hackers to that company. The DeepSeek follows the honorable road of hypersonic missiles and stealth fighters. That algorithm came from nowhere. 

Maybe some part of the code is developed using methods that will not accepted in Europe or other Western countries. The Chinese government uses hackers to steal data that it thinks to benefit it. And there is no way to resist the Chinese government, if and when it wants to participate in some AI project. 

And it's possible. Only the people. Those who are hired by a company. That markets the software take salaries from that work. The entire company can be coulisse. And the project itself is controlled by the People's Liberation Army, PLA. Or, would you download software that the developer is marked as  PLA or "National Police of China"? 

The company's purpose is to make the software or application more attractive. The other thing is this. It's possible. The Chinese government didn't even ask the DeepSeek needed its support. If DeepSeek says "no" that means the end of the business. 

As I wrote before, it's possible. The Chinese government's hackers gave information about DeepSeek. The hacking that happens under the Chinese government's control is suspected of being behind China's fast technological advance. And AI development is at a high level on government's priority. Chinese authorities are interested in the AI's military, intelligence, and surveillance applications. 

That means there is one way in China that developers must follow. That is if the government suggests something the only way to answer is "yes". Other answers cause problems in business. Or, the person who gives that wrong answer will transform to mine worker. 

https://interestingengineering.com/culture/deepseeks-ai-training-cost-billion

Images. InterestingEngineering. 

DeepSeek and its connections.

When you use DeepSeek you know what you use. The program runs on Chinese servers. And it must follow Chinese regulations.  That is the thing in every program. The system tells that it collects data, and that's it. That system is free to use. And that is one of the reasons why we should think about AI very seriously. In the case that somebody does something in China. There is a government behind or with that thing. 

If DeepSeek is freeware. There must be something that its competitors offer. Or Chinese company takes their customers. The question is: Where does the Chinese company behind the Deep Seek get money to pay workers their salaries? The free software doesn't bring money to the company. And that is a good thing to remember. So, DeepSeek is an interesting thing. I tested it. And it was quite an interesting experience. But it won't make images. So, that's why the image above was made using Gemini. 

There is a possibility that the Chinese want to pull rags from below Western AI companies. So, when the Chinese give their AI for free, nobody wants to pay the price. And that takes customers from Western AI companies and ends their research work. The DeepSeek is the tool that makes everything different. It uses powerful and advanced algorithms. 

And it's lightweight. The DeepSeek will make AI development more effective. It doesn't need the most advanced microchips. And that means it's easy to run on a little bit older hardware. Because the DeepSeek doesn't need the most powerful microchips. Is easy to use in countries where the newest microchips are not allowed to transport. That new AI boosts the AI advance on all fronts. The Western companies must have something to offer or their Chinese competitor will take their marketing position. 

The Chinese AI leaves the Western competitors behind.

When we start to use some new product, we must read what it does. Then we must find out what else that product does than its makers tell in commercials. Many times things that are not listed are more important than things that producers of the product tell to customer. The Trojan Horse is the spyware that plays some utility programs. Those programs include a hidden algorithm that dumps the hard disk and net traffic into the net. 

The new and hot thing in the AI business is the Chinese DeepSeek. That artificial intelligence runs on lighter platforms than Western competitors. And that makes it interesting. The fact is that the Chinese AI runs on Chinese servers. Which means that thing has connections with Chinese official policy. DeepSeek has one similarity with the other Chinese innovations. That means the DeepSeek seems to come from nowhere. There was nothing before the publishing. 

And then suddenly the most innovative AI in the world is reality. Sometimes somebody can say that there is suspicion that the Chinese hackers stole the code for the AI, which is similar to other AI's. That thing means that the China took a big step in AI. But the thing that the Chinese seek might be a surveillance tool that can follow and map Chinese citizens. 

The DeepSeek might be only one part of the AI development program in the East. That AI might have the "evil twin" that makes the next-generation hacking tool possible. The advanced AI is a tool that can serve military purposes. The highly advanced algorithms can remove databases from the hard disks. That makes the AI itself is weapon. 

The new road for AI is that it turns very important thing. The AI can generate virus code very fast. And that's why AI can support other systems. But the fact is that. Chinese AI programs are not possible without the government's acceptance. 

The government in China wants to use AI as an intelligence and control tool. There are no limits to the information collection in China if it serves the government's and state interests. AI is a tool that can revolutionize the Chinese control policy and their hacking- or cyber espionage projects. And that is one of the things that we must realize. China and its government know its value and power in the world.  

https://edition.cnn.com/2025/01/27/tech/deepseek-ai-explainer/index.html

https://en.wikipedia.org/wiki/Artificial_intelligence_industry_in_China