Created by an AI assistant Chinese startup DeepSec The app became the number one most-downloaded app in Apple's US App Store over the weekend, sending shock waves through Silicon Valley and sending the price of major tech stocks tumbling. More views than Nvidia $460 billion It was wiped off its market capitalization on Monday, a drop Bloomberg described as “the largest in US stock market history”.
The shake-up stems from an open source model developed by DeepSec called R1, which debuted earlier this month. The company said it rivals the current industry leader: Openai's 01But what shocked the tech industry most was that DeepSec claimed to build its model using only a tiny fraction of the specialized computer chips that AI companies typically need to develop cutting-edge systems. it occurs.
On Monday, DeepSec said it was temporarily limiting new registrations, citing “large-scale malicious attacks” on the company's services, according to a Message On your website.
DeepSeq's R1 model “challenges the notion that Western AI companies have a significant edge over Chinese ones,” says Jack Clark, cofounder of AI startup Anthropic. wrote In your newspaper. Venture capitalist Marc Andreessen called This is “AI’s Sputnik moment.”
Cheng Lu, a research scientist at Openai, said DeepSec's chatbot demonstrated impressive Chinese conversational skills. “This is the first time I can feel the beauty of the Chinese language created by a chatbot,” he said in an x post On Sunday.
DeepSec's AI assistant is currently available for free and comes with three main functions. First, users can ask the chatbot questions and get direct answers. For example, when Wired asked for recipe ideas incorporating pomegranate seeds, DeepSec's chatbot quickly provided a list of 15 options ranging from yogurt parfaits to “Middle Eastern-inspired” rice pilaf, but it Did not cite any specific chefs or cuisines.
DeepSec's app also has a search mode that surfaces answers from the Internet. When Wired asked “What are some important news stories today?” DeepSec's chatbot cited the Israel-Hamas ceasefire and linked to several Western news outlets such as BBC News, but not all stories were relevant to the topic. Ironically, there was a New York Times story about the impact of DeepSec on the stock market.
Finally, there's a “DeepThink” mode that allows users to tap into DeepSeq's R1 model, which was built on the company's existing V3 modelThe difference between the two is that R1 has so-called “reasoning” capabilities that allow it to explain step by step how it reached its conclusions. For example, when asked “What are the most important historical events of the 20th century?” DeepSec initially responded with a lengthy answer that began with several broad questions.
“It's been a hundred years, so a lot has happened,” read part of its answer. “I should probably break it down by decades or major topics like wars, political change, technological advancements, social movements, etc.” DeepSec's chatbot then cited World War II, the Cold War, and the Holocaust.
But before R1 could finish his answer, the entire answer disappeared and was replaced by a message that read “Sorry, I'm not sure how to approach this type of question yet. Let's move on to math, coding.” And chat about logic problems instead” Many experts and early adopters! Have noted Like other tech platforms operating in China, it appears DeepSec mass sensor Sensitive topic by Chinese Communist Party
But despite these limitations, DeepSec's free chatbot could pose a serious threat to competitors like OpenAI, which charges $20 per month for access to its most powerful AI models. Unlike its Chinese counterpart, Openai does not disclose the underlying “weights” of its models, which determine how the AI processes information. It has also refused to make a public statement.”chain of thought“Built by its own logic model.