Just admit it—you’ve used Artificial Intelligence (AI) in some way, whether or not you are proud to admit it. Whether for academics, entertainment, or other purposes, the rise of AI chatbots has made society increasingly reliant on their use. One of the latest developments is DeepSeek R1, a Chinese- developed AI bot. Despite its recent introduction to American app stores, its release caused major US technology stocks to tank, and it has already been banned on governmental devices in Australia, South Korea, Italy, and even here in New York.
Just one week after it was released into American app stores, DeepSeek reached No. 1 in the Apple App store, surpassing ChatGPT-4 for the top spot. Its creation only took about 5.6 million dollars to develop, which is a fraction of the amount it took for American companies such as OpenAI to develop their AI models. On top of this, developers of DeepSeek had limited access to technologies that aided its development compared to those available in the US.
DeepSeek was originally developed in 2023 by Liang Wenfeng. It functions similarly to American models, such as ChatGPT, yet is able to perform these tasks with less energy and memory. DeepSeek is embedded with a “reasoning” approach and a “Mixture-of-Experts” model. This means that when inputted with a prompt, cues to answer the prompt will only be signalled to “experts” that know the most about that specific task. This allows DeepSeek to carry out its tasks in an efficient manner, limiting the amount of energy exerted. However, ChatGPT is a transformer model that answers its prompts all at once, as if it grabs information for all possible sources it can find, which uses a lot more energy. DeepSeek, being more energy efficient, was showcased as an extraordinary development for AI’s future. For example, it could reduce the consumption of natural resources used to cool AI data centers, helping to preserve the environment.
While DeepSeek’s efficiency is a large benefit, there are concerns with how users’ information is being stored in its software, and how that information could be directly sent to the Chinese government. Additionally, researchers have found that the software may be subject to censorship by the Chinese government when prompted about violent instances in history that occurred in China, such as the Tiananmen Square massacre in 1989. This leaves one to question how this technology may be censoring specific information it provides. Concerns about how data is stored and used are a major reason why many countries and states are banning the platform from government devices.
DeepSeek’s development has proven to be a major change in the future of AI. It has the potential to be an example to the development of future chatbots in its reasoning model and energy conservation. Still, its ability to filter and collect personal information from users makes its use unjustifiable. It is important to keep up with how AI is currently being developed, as society begins to incorporate and innovate technology dependent on the world of AI. Just as ChatGPT has become a source many have grown to rely on, DeepSeek can be—and frankly, is—as well. That’s why staying cautious and aware of how it is being developed is essential, both now and in the future.
How does DeepSeek’s development impact the world of AI? NBC News
Be First to Comment