AI Blog: DeepSeek Tour Part 2

Welcome back to our weekly AI blog.  Today, we start taking a more in depth look at the AI tool that has recently captured so much public attention.

Our previous blog gave a quick review of the working theory of Artificial Intelligence (AI) and compared the former industry leader, OpenAI’s ChatGPT and the rising star DeepSeek in terms of the price, accuracy and speed.  This time, we will introduce how using this tool might make your life easier.  Let’s begin with DeepSeek. 

DeepSeek

“DeepSeek-R1 is now live and open source, rivalling OpenAI's Model o1.  Available on web, app, and API.”

As mentioned last time, DeepSeek is an open-source AI chatbot accessible at https://www.deepseek.com/.  You may also download its application by scanning the QR code or going directly to the App Store.  You may already be signed in to a Google account in your web browser, but if not, you will be prompted to sign in:

Either log in or sign up for a free account and start exploring; once you access the user interface (UI), you will find it familiar, as the DeepSeek UI is nearly identical to other AI tools including ChatGPT, Bard and Gemini.  The ‘Open Sidebar’ contains the chat history of the last 7 days, 30 days, and even earlier, allowing you to revisit questions and answers or continue previous conversations.  Each conversation may be renamed for easy reference.

The default model is DeepSeek-V3, which achieved a significant breakthrough in inference speed, supports multiple languages.  It excels in English, code, mathematics and Chinese.

At the bottom of the prompt box, you’ll find three buttons: DeepThink (R1), Search and an attachment icon.  These allow users to switch to the advanced deep think model (comparable to ChatGPT’s o1 model), search the website to generate the latest data and upload files or images.

DeepSeek is an advanced Large Language Model (LLM).  You can ask it any questions on any topic.

Join us next time to explore some of its most common applications.