PUBLISHED
February 02, 2025
KARACHI:
Could a newcomer disrupt the established order in the AI world? That’s the question on everyone’s mind as China’s DeepSeek AI model makes waves with its claim to deliver performance comparable to giants like OpenAI, Google, and Meta, but at a fraction of the cost. This unexpected development has sent tremors through the industry, triggering a reassessment of AI investments and causing significant volatility in the stock market. What makes DeepSeek so disruptive, and what does its emergence mean for the future of AI?
A market in freefall
The financial world reacted swiftly to DeepSeek’s arrival, with the total value of AI companies plummeting by over $2.5 billion in 24 hours. This decline coincides with DeepSeek’s rise and has impacted various sectors of the AI market. Leading AI agents like AIXBT, ai16z framework, Virtuals AI agent creation platform, and GRIFFAIN have all experienced significant drops, ranging from 14% to 25%. Even the crypto market has felt the impact, with AI meme coins like Fartcoin and tokens like Near Protocol, Internet Computer, Render, and Artificial Superintelligence Alliance all experiencing declines.
The stock market has been particularly volatile, with US stocks erasing more than $1 trillion in market cap on Monday. Chipmakers, essential for AI infrastructure, have been among the hardest hit. Nvidia, the dominant player in this sector, saw its stock plunge by 16.9%, wiping out $589 billion from its market value. This represents the largest single-day loss of market cap in history, a stark illustration of DeepSeek’s disruptive potential. Other major chipmakers like Broadcom and Taiwan Semiconductor Manufacturing Company have also suffered substantial declines, with both experiencing a 17% drop in their stock value. ASML, another key player in the chipmaking industry, has also been affected, with its shares falling by 6%.
The impact extends beyond the US, with European tech stocks also feeling the pressure. The pan-European STOXX 600 index slid by 3.4%, and chip equipment maker ASML experienced a 7% decline. This global reaction highlights the widespread concern about DeepSeek’s potential to disrupt the established order in the AI industry. This widespread market reaction underscores the uncertainty and anxiety surrounding DeepSeek’s emergence. Investors are clearly concerned about the potential implications for established players and the future direction of AI investments.
DeepSeek’s potential long-term impact
Before delving into the reasons behind the market’s reaction, let’s explore the potential long-term implications of DeepSeek’s emergence. Several key scenarios are emerging.
DeepSeek’s open-source approach and cost-efficiency could lower the barriers to entry for AI development, fostering innovation and competition. This could lead to a more diverse and dynamic AI ecosystem, with new players emerging and challenging the dominance of established giants.
As AI becomes more affordable, we may see a surge in its adoption across various industries. This could lead to increased productivity, automation of tasks, and the development of new solutions to complex problems. This potential extends beyond the tech sector, offering opportunities for non-tech companies to leverage AI for increased productivity and growth.
Investors may shift their focus from AI infrastructure to AI applications and services. This could lead to increased funding for startups and companies developing innovative AI solutions.
DeepSeek’s success could intensify competition in the AI market, driving further innovation and potentially leading to price wars. This could benefit consumers and businesses alike, as AI products and services become more affordable and accessible.
DeepSeek’s efficiency has the potential to disrupt the US power market by reducing energy consumption for AI applications. This could lead to a reassessment of long-term electricity demand projections and the sustainability of current data center growth rates.
DeepSeek’s emergence has also highlighted the growing competition between the US and China in the AI domain. This could lead to increased investment in AI research and development on both sides, as well as potential trade tensions and restrictions.
These potential long-term impacts highlight the transformative potential of DeepSeek’s arrival. It’s not just about a new AI model; it’s about a potential paradigm shift in how AI is developed, accessed, and utilised across various sectors.
The changing dynamics of the AI market
DeepSeek’s emergence has triggered a fundamental shift in the dynamics of the AI market. The focus is moving away from infrastructure spending and towards the development of practical AI applications. As AI models become more affordable, the emphasis is shifting to what can be built with AI, rather than how much compute power can be amassed.
This shift is occurring within a broader context of rapid evolution in the AI market. The adaptive AI market, for example, is estimated to grow at a CAGR of 44.71% during 2024-2029. This growth is driven by the increasing demand for AI systems with improved capabilities and efficient performance across multiple industries.
DeepSeek’s emergence could further accelerate this shift towards applications by intensifying the “Jevons Paradox” in the AI market. This paradox, named after economist William Jevons, suggests that technological improvements in efficiency often lead to increased overall usage rather than reduced consumption. In the context of AI, cheaper and more efficient models like DeepSeek’s could spark widespread adoption, driving demand for AI infrastructure and applications in the long run.
Why the market is panicking
At the heart of the market’s reaction lies DeepSeek’s remarkable cost-efficiency. DeepSeek charges a mere 14 cents per million input tokens, significantly undercutting OpenAI’s GPT-4, which costs $15 per million input tokens. This stark difference has forced investors to re-evaluate the financial models of established AI companies, raising concerns about their ability to compete in a landscape where efficiency reigns supreme.
DeepSeek’s success stems from a combination of factors, including its reliance on open-source technology and innovative training methods. Unlike proprietary models developed by companies like OpenAI, Google, and Meta, DeepSeek leverages open-source tools and allows for modification and access by a wider range of players. This open-source approach has the potential to reshape the AI ecosystem by fostering collaboration, knowledge sharing, and a more decentralised development landscape.
DeepSeek has also demonstrated that it can achieve comparable performance with fewer resources, challenging the assumption that massive infrastructure spending is essential for AI dominance. This has been achieved through innovative techniques, such as skipping supervised fine-tuning in favour of pure reinforcement learning and using “Aha moments” as pivot tokens. DeepSeek also utilises “inference-time computing,” which means it activates only the most relevant portions of its model for each query, saving money and computational power.
These factors combined have enabled DeepSeek to build powerful AI without enormous hardware outlays, prompting investors to question the value of AI infrastructure and hardware. This has led to concerns about the future profitability of companies like Nvidia, which have heavily invested in specialised AI chips and data centers.
Concerns and challenges
While DeepSeek’s emergence presents exciting possibilities, it also raises important concerns and challenges.
DeepSeek’s data collection practices have raised concerns about user privacy. The company stores data from users, including their date of birth, keystrokes, text or audio inputs, uploaded files, chat history, and other data, on servers located in China. This raises questions about the security and potential misuse of this data, particularly given the different data privacy regulations in China compared to other countries.
As with any AI model, DeepSeek is susceptible to bias and inaccuracies. This is particularly concerning given its open-source nature, which allows for modification and potential manipulation of the model. Bias in AI systems can lead to discriminatory outcomes, while inaccuracies can spread misinformation and erode trust in AI technologies.
These concerns highlight the need for robust governance and ethical frameworks to guide the development and deployment of AI, particularly as it becomes more accessible and influential in various aspects of our lives.
Navigating the DeepSeek disruption
DeepSeek’s arrival has undoubtedly shaken the AI market, forcing a reassessment of existing strategies and investment priorities. While the short-term impact has been negative for many established players, the long-term implications are more nuanced. DeepSeek’s efficiency and open-source approach have the potential to democratise AI, accelerate its adoption, and drive further innovation.
The AI industry is likely to adapt to the challenges posed by DeepSeek in several ways. We may see increased investment in research and development to improve the efficiency and capabilities of existing AI models. Companies may also focus on developing specialised AI applications and services that cater to specific industry needs, rather than relying solely on general-purpose models.
The potential long-term benefits of this disruption include increased accessibility to AI, greater innovation, and a more competitive market that could lead to lower prices and improved quality for consumers. However, there are also risks, such as the potential for increased bias and inaccuracies in AI systems, as well as the need to address data privacy concerns.
Governments and regulatory bodies will play a crucial role in shaping the future of AI in this new landscape. They will need to establish clear guidelines and regulations to ensure the responsible development and deployment of AI, while also fostering innovation and competition.
For investors and businesses navigating this changing landscape, diversification and adaptability will be key. Investing in a range of AI companies, technologies, and applications can help mitigate risk. It’s also crucial to stay informed about the latest developments in AI and be prepared to adjust strategies as the market evolves.
DeepSeek’s emergence is a reminder that the AI landscape is constantly evolving. While there may be short-term volatility and uncertainty, the long-term potential of AI remains immense. By embracing innovation, addressing challenges responsibly, and adapting to the changing dynamics of the market, we can harness the power of AI to create a more productive, efficient, and equitable future.
Ali Asad Sabir is working as project manager at Mahbub ul Haq research centre at Lahore University of Management Sciences (LUMS)
All facts and information are the sole responsibility of the author
