Venture capitalists have poured significant funds into AI startups like OpenAI and Anthropic. However, the rise of the Chinese AI startup DeepSeek has called this funding frenzy into question. DeepSeek has developed a powerful AI model with far less money than most tech experts believed possible, challenging many assumptions about the rapid evolution of technology.
Jordan Jacobs, an investor at Radical Ventures, has spent recent days fielding calls from investors eager to learn more about DeepSeek, which topped app store charts over the weekend. To ease their concerns, Jacobs explained that Radical Ventures has long invested in more efficient AI models, similar to what DeepSeek has achieved. He emphasized the importance of focusing on companies that are building real businesses rather than those chasing science fiction.
Since last week, tech giants like Nvidia, Google, and Meta have faced a barrage of questions regarding DeepSeek. The startup’s success has not only disrupted established notions about AI but has also impacted the venture capital sector that has heavily invested billions in AI startups. This shift has raised concerns about the sustainability of the funding model that has dominated the industry.
Over the past two years, venture capital firms have engaged in a funding frenzy, pouring over $155 billion into AI startups between 2023 and 2024, according to PitchBook. Notably, OpenAI and Anthropic raised $24 billion and $16 billion, respectively, with the goal of developing human-level AI. OpenAI’s valuation reached $157 billion, surpassing that of Pfizer and Citigroup, while Anthropic’s valuation hit $20 billion.
DeepSeek’s recent achievements have put this funding frenzy into question. If a Chinese newcomer can create an app as powerful as OpenAI’s ChatGPT or Anthropic’s Claude chatbot with minimal investment, it raises the question of why these established companies needed to raise such vast sums of money. This situation has led to a reevaluation of the investment landscape in AI.
Matt Turck, an investor at FirstMark Capital, noted that the current climate is not favorable for some AI companies, especially given the conversation around the need for ever-increasing scale to achieve the best models. However, he added that AI companies will ultimately require funding, computing power, and infrastructure to serve their customers effectively. This ongoing debate highlights the complexities of investing in AI technology.
As the excitement surrounding DeepSeek continues, investors who previously avoided foundational model companies like OpenAI and Anthropic are seizing the moment to share their opinions. Eric Vishria, a Benchmark investor, expressed on social media that he believes foundational models are “the fastest depreciating asset in human history.” This sentiment reflects a growing skepticism about the long-term viability of such investments.
Despite the challenges posed by DeepSeek’s success, investors who backed foundational model companies are defending their investments. Gavin Baker, an investor at Atreides Management, expressed confidence in his investment in Elon Musk’s AI startup, X.ai, citing its unique data source from the social network X. He believes that having access to proprietary data will position X.ai favorably in the competitive landscape.
In conclusion, the emergence of DeepSeek has sparked a significant shift in the AI investment landscape, prompting both excitement and concern among investors. As startups begin to explore new techniques for developing AI models that are cheaper to train and maintain, the potential for increased competition and innovation in the sector is becoming apparent. The dynamic nature of the AI space suggests that investors should brace for more surprises in the weeks to come.
