In brief
While testing DeepSeek's AI chatbot from within China, the chatbot revealed a dual nature: sometimes evading censorship and at other times self-censoring. This raises questions about how this technology could influence global narratives if adopted internationally.
As the global tech landscape evolves, understanding the intricacies of AI and censorship becomes increasingly important, particularly when examining products painted with the brush of geopolitical implications. With a focus on DeepSeek’s A.I. chatbot, I initiated an experiment to uncover how this technology interacts with China's stringent internet censorship characterized by the Great Firewall. Having registered for the service using a Chinese phone number and internet connection, I found myself in a unique position to explore its capabilities under restrictive conditions. Surprisingly, DeepSeek often delivered answers that skirted around the strict confines of typical censorship seen on popular Chinese platforms. In various instances, not only did the AI manage to provide insights labeling topics digitally taboo, but its "thinking" feature also allowed it to articulate answers before self-censoring. This feature proved pivotal as it offered users a glimpse into the unfiltered information that typically eludes mainstream Chinese social media.Moreover, my attempts to replicate the same inquiries using methods to bypass the Great Firewall yielded similar responses, suggesting that DeepSeek aims for consistency in user experience, whether domestic or international. However, this also raises concerns regarding the potential global spread of narratives that align with the Chinese government's perspective, should DeepSeek gain traction beyond China's borders.As the world becomes more receptive to AI developments, the implications of tools like DeepSeek on international discourse remain critical to monitor for the sake of information integrity and cultural understanding in an increasingly interconnected world.
