Is DeepSeek Better Than ChatGPT?

While majority of mainstream media is backing DeepSeek, user experiences vary depending on the type of queries

Within just a few days of its launch, DeepSeek’s latest version has rattled the global AI and tech industry. US tech and AI stocks have seen a fall of more than $1 trillion in their valuation. Across most media platforms, one can see reports on how DeepSeek is better than ChatGPT. Let us do a quick evaluation to understand if DeepSeek is really better than ChatGPT.

DeepSeek Vs. ChatGPT – Key Differences

Open source – One of the things that make DeepSeek more accessible is that it is an open-source platform. It implies that developers across the globe can contribute to enhance the code. This will ensure that DeepSeek continues to improve with every new version. In comparison, ChatGPT is developed and maintained exclusively by OpenAI. The training models and underlying code of ChatGPT is proprietary. However, being proprietary is not necessarily bad. It is just that DeepSeek’s open-source approach will attract more developers who may want to make the platform better.

Affordability – For organizations that may want to benefit from AI, DeepSeek can be a more affordable option. DeepSeek has been developed using fewer resources. For example, most of the leading AI firms use supercomputers to train their chatbots. They have used up to 16,000 integrated circuits for such tasks. In comparison, DeepSeek V3 was developed using just around 2,000 specialized computer chips. The training time was just 55 days and the overall cost was US$5.58 million. That is around one-tenth of the costs involved in developing AI tech by other big companies like Meta.

User experiences – As of now, the digital realm is flooded with numerous reports from mainstream media. As a result, actual user experiences are not getting the attention they deserve. Ultimately, it will be the user experiences that will determine which generative AI platform is better. It is easy to influence mainstream media, but it can be done only for a specific period of time. Some actual user experiences are now starting to get highlighted. And these reveal mixed reactions – some in favor of DeepSeek and others voting for ChatGPT.

Questioning skills – What type of response one gets from an AI platform also depends on the querying skills of the user. While AI platforms are designed to understand badly quoted questions, it may still impact the response generated. Learning to ask the right questions in an easy-to-understand language can help get better results from generative AI platforms such as DeepSeek and ChatGPT. It implies that deciding which is the better AI platform also depends on who is asking the questions.

Privacy and censorship – DeepSeek is developed by a China-based company, something that has raised concerns about privacy and censorship. For example, DeepSeek does not respond to sensitive topics such as alleged human rights violations in China, Uyghurs and 1989 Tiananmen Square protests and massacre. Discussions are also terminated if one questions any of the core socialist values. It is also being claimed that DeepSeek could be collecting personal data and transferring it to servers in China. This data can be misused for spreading disinformation, carrying out surveillance and developing cyberweapons.

From the above, it is apparent that one needs more time to assess whether DeepSeek is better than ChatGPT. As of now, the media deluge is making it look like a one-sided game. One will get a better idea when the focus shifts to actual user experiences.