Last week, the Chinese AI-chatbot app DeepSeek was released, causing havoc in US markets and casting doubt on America’s continued superiority in AI. The BBC examines the operation of the app.
Though it tends to be too talkative, DeepSeek feels and looks like any other chatbot.
Similar to Google’s Gemini or OpenAI’s ChatGPT, you launch the app (or website) and ask it any question, and it tries its best to respond.
It provides lengthy responses and refrains from voicing an opinion, even while one is explicitly requested.
Frequently, the chatbot starts off by stating that the issue is “highly subjective”—whether it’s politics (is Donald Trump a good US president?) or something else entirely. or soft drinks (Coke or Pepsi, which tastes better?).
It weighed the advantages and disadvantages of each, but it wouldn’t even promise that it was superior to ChatGPT, OpenAI’s competing artificial intelligence (AI) assistant. ChatGPT performed the exact same tasks and even employed language that was strikingly similar.
Although the app appears to have access to up-to-date data, including the current date, the website version does not, according to DeepSeek, which claims that it was trained on data up to October 2023.
The attempt to prevent the chatbot from blazing out false information that has been blasted over the internet in real time is likely similar to previous iterations of ChatGPT.
ICYMT: Appointments Committee Summons Vormawor Over Bribery Allegations
It can react quite quickly, but right now it’s complaining about how many people are trying it out because it’s so popular.
However, DeepSeek differs greatly from its US competitor in one aspect: it filters out queries pertaining to topics that are prohibited in China.
