Looking on the bright side of AI: seeking the good in DeepSeek

At the end of January 2025, a friend of mine from China sent me a note on the social messenger service WeChat: “You need to read this article! Read it carefully!” It was an article about DeepSeek, the Chinese AI-powered chatbot, on a WeChat public media account.

Unfortunately, two years ago WeChat sent out a clause about privacy that I didn’t choose to agree to, so since then, I’ve never been able to open these public accounts. I am using WeChat more and more rarely.

So, my friends have to convert WeChat articles into PDFs or forward them to another social media platform before I can see them. My friend couldn’t wait for my response and told me bluntly, “My conclusion is that in time, AI can completely replace human writing. Some of the things we humans are most proud of don’t really matter because we lost them long ago without realizing it.”

I knew he was serious. And so I began to seriously look at DeepSeek. I had always thought that AI couldn’t replace artistic creations—those AI-generated texts, music, drawings and videos tend to repulse people after being watched a few times.

I was in a trance, shocked by its “creativity.”

But this time, honestly speaking, after reading the wuxia novels (Chinese stories about knights doing kung fu) and a few modern poems generated by DeepSeek, as a die-hard literature fan I was in a trance, shocked by its “creativity.”

Obviously, its ability to handle the Chinese context is much more robust than ChatGPT, just as WeChat is better than Facebook for the Chinese context.

DeepSeek, like WeChat, has the limitations due to its Chinese characteristics. Unsurprisingly, it automatically avoids all the sensitive topics you can think of, like politics, history and so on.

Interestingly, its advice on attending house churches is not to avoid them, but to keep a low profile. And it recognizes that this is a response to “policies at the top and countermeasures below.” Apparently, China’s AI also has “Chinese” survival wisdom.

We know that DeepSeek has a good understanding of the taboo topics, but it is not allowed to tell us. It has the same limitations that Chinese people face every day. Some people are worried about these limitations, but if China opens up one day, none of this will be a problem—the capacity is there. 

The limitations are political, not technological.

The limitations are political, not technological. I’m also not too worried about whether it will replace human writing in the future because how many people will be reading by then? I do have some concerns, though, about security and the privacy of user information, and the AI generated false information.

Yet I’m even more worried about the fact that since 2010, search engines, social media, and algorithmic feeds, which are all simpler technologies than DeepSeek, have had a huge impact on us that we have yet to deeply reflect on. We have all, including Christians, become digitally captive, both in China and elsewhere. 

The sheer volume of information has exceeded our human capacity to receive it, and in the face of false information, we are unable or unwilling to distinguish truth from lies. We live in a world of information that is opposed to each other, and our consensus on the truth, the good, and the beautiful is becoming thinner and thinner. 

In 2017, we conducted a study of Christian content on WeChat and concluded that bad money drives out good money. The “bad money” here is a reference to what’s counterfeit—so the abundance of falsehood expels the truth, which expels the consensus on the good, and also expels beauty step by step.

One of the things that makes DeepSeek different from similar products is that it shares with you its processing or thinking process, i.e. logic. This process itself shows its ability to do exactly what is lacking among people today who are used to being consumers who are fed information uncritically.

It trains you—helping you relearn critical thinking.

DeepSeek talks to people in a seemingly natural way, interacting less like a secretary and more like a brainstorming session with a colleague or partner. As you train it, without realizing it, it trains you—helping you relearn critical thinking and discretion from the very software we created.

This, in turn, is where I feel hopeful. We all need to learn to think critically, apply logic, and organize our own thoughts. And Christians in particular need to “deep seek” more for truth, goodness, and beauty.

As artificial intelligence rapidly advances, its impact on human creativity, information consumption, and even spiritual life becomes an urgent topic of discussion. DeepSeek, a China-developed AI chatbot, not only highlights the technological strides in natural language processing but also reflects the unique challenges of navigating digital spaces under China’s specific constraints.

For Christians, particularly those in house church communities, AI’s role in shaping access to and interpretation of information is both a cautionary and an intriguing development. Realizing this is timely and important as it raises critical questions about AI’s influence on truth-seeking and the future of human thought, encouraging deeper reflection on how we engage with technology in faith and society.

Originally published in English by China Source. Republished with permission.

Pastor Jerry An has worked in media ministry since 2001, and now serves as the Chinese Team Leader at ReFrame Ministries (formerly Back to God Ministries International). Under his vision and leadership, the Chinese language ministry of ReFrame has become a pioneer, think tank, and partner in new media ministry. Pastor Jerry regularly trains Chinese church leaders worldwide in developing ways of using new media to advance the kingdom. He is also a publisher, producer, radio host, and writer.