In a world where artificial intelligence is increasingly integrating itself into everyday life, one platform has skyrocketed to the top. DeepSeek is an innovative AI tool that has been starting to rise in the App Store, beating Chat GPT and Google AI in the rankings despite operating on a significantly lower budget.

Founded by Liang Wenfeng, the application is designed to process and generate human-like text with remarkable efficiency. These models are available for anyone to use, modify, and distribute. DeepSeek employs an expert foundation that separates the model to activate the excerpts needed for specific tasks. For example, DeepSeek has a section specifically for questions about math, English, or science. Ultimately, this app has caused a commotion in the AI world, leading the US tech stocks to sink.

“It’s pretty surprising but to me, AI can be helpful to students and even other people, but only if it’s used wisely can it be helpful to many students,” math teacher Maria Terada said. “For example, I’ve seen other students that use it the wrong way and let it do the work for them, and now they do not understand the lesson because AI did the work for them. It’s interesting to see that there’s more advanced AI like DeepSeek out there.”

Despite the economic turbulence, DeepSeek has quickly gained a loyal user base, particularly among students and programmers who admire its ability to analyze large amounts of data and assist with complex tasks.

“It has helped me with some programming problems by being able to upload attachments of my project, and it has been able to read over all my code and provide valid feedback and enhance my ideas,” Ethan Mikes said. “In terms of some school work, I can upload my previous assignments and ask it to build off of the directions of my current work.”

DeepSeek has grown larger than other AI models because it uses only a fraction of its parameters in each interaction. For instance, since each submodel is activated only when the particular knowledge is needed, DeepSeek has been able to use fewer of its parameters. According to a technical report published by the developers, this app has 671 billion parameters; however, it only uses around 37 billion at a time.

“I’ve heard a couple of things about DeepSeek and I honestly believe that it is better than other US AI tools,” junior Gurnoor Gill said. “Apparently it uses way less storage and is able to more accurately give information about the subject that you ask about.”

Even though this application has been used as a learning tool for many people, there have been some downsides to this app. With the potential concern of DeepSeek stealing information or privacy, users have been cautious about the risk of this. Unlike open AI models, which are subject to U.S. regulations, DeepSeek is under Chinese laws which raise concerns about how user data is stored and utilized.

“I’m assuming it operates like most AI and can, you know, be beneficial in that realm, but it does sound like it is censored, as is all AI,” English 11 teacher Jacob Hill said. “However, there have been some things about the app that I find a little weird, especially when asking about things like the Chinese Government, it completely censors all of that stuff.”

Regulators are already discussing the possibility of exhibiting the use of DeepSeek by the executive agencies. Presenting the No DeepSeek on Government Device Act, Rep. Josh Gottheimer hopes to prevent the usage of this AI app. Furthermore, the Office of the Texas Governor announced a ban on Chinese AI on January 31st, stating that they would not allow AI or social media apps affiliated with the People’s Republic of China or the Chinese Communist Party on any government-issued devices.

“If DeepSeek is getting banned due to concerns that the app might be stealing information from users or privacy from them, I don’t think it should be banned,” Hill said “I see so many corporations already doing that through apps that we use, and with even just existing in the internet there’s a risk of it. So I think banning that one would open up a case of banning a lot of other things. If there is a way to limit the data that is taken from users, then that is okay, but if it is inherently stealing data then I understand the cause for banning.”