The world of AI is in a frenzy over DeepSeek's Liang Wenfeng, a rising star in the field. But is he a genius or a disruptor?
Nature, a renowned British journal, has named Liang Wenfeng, founder and CEO of DeepSeek, among the top 10 influential figures in science for 2025. This recognition comes as no surprise, given the 40-year-old's groundbreaking work in artificial intelligence.
Here's the twist: DeepSeek's AI models, particularly the reasoning model DeepSeek-R1, challenged the notion of US dominance in AI. The model's release in January caused a stir, revealing that the US might not be as far ahead in AI development as previously believed. And this is where it gets controversial—DeepSeek's technology sent shockwaves through the US stock market.
On January 27, a massive sell-off triggered by DeepSeek's models wiped out a staggering US$1 trillion in tech stocks, with Nvidia taking a US$600 billion hit. But that's not all. DeepSeek's models, R1 and V3, released in December, performed remarkably well, rivaling OpenAI's offerings at a much lower training cost. This achievement questioned the high valuations of US semiconductor and AI companies, as their training costs were significantly higher.
For instance, Meta Platforms' Llama 3 405B model had training costs more than ten times greater than DeepSeek's, according to Nature. DeepSeek's open-source models have become a game-changer, allowing researchers to adapt algorithms freely, and inspiring other companies to embrace open-source practices.
Adina Yakefu, a researcher at AI developer platform Hugging Face, praised DeepSeek's impact, stating that it has been 'hugely influential' in the AI community. But the question remains: Is Liang Wenfeng a hero or a disruptor in the eyes of the AI industry? What do you think? Share your thoughts on this fascinating development!
