Meta AI Chief And Others Praise DeepSeek’s Open-Source R1 Model
Contents
The United States is home to some of the biggest AI companies like Google, Meta, OpenAI, Anthropic, and more. No wonder, the country has been one of the undisputed leaders in the artificial intelligence (AI) industry. But it seems DeepSeek, a Chinese AI startup, is looking to change that narrative with its open-source large language model – DeepSeek-R1. DeepSeek’s new open-source reasoning model has reportedly taken Silicon Valley by storm as it outperformed existing models from OpenAI, Meta, and Anthropic in third-party benchmarks.
Meta’s chief AI scientist praises DeepSeek’s new open-source reasoning model
DeepSeek, in its release notes, mentions that the DeepSeek-R1 reasoning model uses reinforcement learning (RL) and multi-stage training to improve its reasoning ability. The Chinese AI startup also claims that the performance of its open-source model “is comparable” to OpenAI o1 on reasoning tasks.
DeepSeek’s achievement has been lauded by many investors and popular figures in the AI industry. Meta’s chief AI scientist, Yann LeCun, is one of them. It’s no surprise that many would think that a China-based startup’s success is an immediate threat to the US in AI. However, LeCun has a different point of view.
In a recent post on Threads, he notes, “DeepSeek has profited from open research and open source.” He further adds “They came up with new ideas and built them on top of other people’s work. Because their work is published and open source, everyone can profit from it. That is the power of open research and open source.”
Among the names is Venture capitalist Marc Andreessen, who recently took to X (formerly Twitter) to praise DeepSeek’s achievement. He noted it is “one of the most amazing and impressive breakthroughs I’ve ever seen.”
Meta announced its AI spending for 2025 days after DeepSeek released the R1 model
Meta, on the other hand, has also announced that it is planning to spend over $60 billion this year. Mark Zuckerberg has always been in favor of open-source models. No wonder, he is shelling out a big chunk towards strengthening its position against the likes of OpenAI and Google.
More recently, a leaked internal memo also indicated that Zuckerberg is planning to lay off low-performing staff. So, it’s safe to say that Meta is quite serious about the goals it is planning to achieve this year.
