AI and democracy.

"AI bias is real. ChatGPT favors left-leaning views, raising concerns about fairness, democracy, and free speech. Researchers urge transparency and safeguards before it’s too late. Credit: SciTechDaily.com" (ScitechDaily, Behind the Code: Unmasking AI’s Hidden Political Bias)

Chinese philosopher Mencius said that people should use criticism when they read texts that Confucius made. The reason for that was that. Confucious was the society class that was just below the leader. His position in society affected Confucius and his texts. So, that's why. Confucius was somehow authoritarian and also conservative. There are many ways to write those names and the source determines how those old Chinese names should be written. 

AI is not as neutral as we wish. The new AI called DeepSeek censors its answers if they include things like the Tiananmen Square. Another thing is that the AI is not allowed to give some answers that could tell you something about how it uses the data that it gets from searches and answers. Another thing is that the AI might not be neutral because its algorithms can use the page rank to select sources. 

This is what makes AI untrustable in some cases. The AI does not think as we think. It just selects sources using some parameters. The problem is that the financial actors can raise their page ranking by buying a place in the search results list. The same things that work with commercial materials work with political parties. 

Same way. Political actors can buy a rise in their ideological material in page ranking. Another way that the actors can raise their material is to make a bot that gives automatic clicks to the home pages. The command is easy to find from the net. It can be used in the homepage group. That sends automatic clicks to some targeted homepages. The command is made for cases where the homepages like web services are transferred to another server or under some other web address. 

The main problem is that the AI is a multipurpose tool. When we make algorithms that give us very interesting data, we face problems where somebody misuses those things. The AI that uses only, let's say, about five first homepages that the search engine gives can cause an effect that the homepage rises in page ranking. That also affects other searches where the homepage is included. And that can cause a situation where the AIs start to raise some homepages. That causes a situation where there are fewer homepages. That is involved in the searches. 

And that causes situations where information that people use turns unilaterally. So, the net follows the route of other news agencies. The unilateral and centralized information sources make news and other information vulnerable to political and financial actors. If, there is only one news agency that gives its owners unlimited power to information. Another thing is that the social media follows that route. 

The "alternative media" where people can share their opinions is owned by somebody. And that "somebody" gives rules. About what is tolerated and what is not. That makes this thing problematic. AI, social media, and other Internet require new types of transparency. People should have information about the things that are behind some systems or applications. The information of the owner of the system can help people form an image of the purpose of the motives behind some applications. 

https://scitechdaily.com/behind-the-code-unmasking-ais-hidden-political-bias/

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

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