MillenniumPost
Big Story

Chinese AI DeepSeek disrupts global tech

Chinese AI DeepSeek disrupts global tech
X

New Delhi/New York: A Chinese artificial intelligence startup has sent shockwaves through global technology markets, causing one of the largest single-day market value losses in history and raising questions about the massive investments in AI by US tech giants.

DeepSeek, a Hangzhou-based research lab, has emerged as a significant disruptor in the AI industry with its latest offerings: DeepSeek V3, a 671 billion parameter ‘mixture of experts’ model, and DeepSeek R1, an advanced reasoning model. The startup’s AI model quickly rose to become the most downloaded app on Apple’s US App Store, marking a notable achievement for a Chinese-developed chatbot in the American market.

The impact of DeepSeek’s emergence was immediately felt in the stock market. Nvidia, the leading AI chip manufacturer, experienced a historic $593 billion loss in market value on Monday—the largest single-day loss for any company. The broader technology sector also suffered, with AI-linked companies collectively losing more than $1 trillion in market value. The Nasdaq 100 fell 3 per cent, its steepest decline in six weeks, while the Philadelphia semiconductor index recorded its largest drop since March 2020.

However, on Tuesday, US markets stabilised following a wipeout a day earlier. Many of Monday’s big losers clawed back some of their losses, including Nvidia, which rose 3.4 per cent in premarket after tumbling nearly 17 per cent a day earlier.

The world’s 500 richest individuals collectively lost $108 billion on Monday. The selloff sent major indices plunging, with the Nasdaq Composite Index dropping 3.1 per cent and the S&P 500 falling 1.5 per cent, according to the Bloomberg Billionaires Index.

Tech billionaires, particularly those tied to artificial intelligence, bore the brunt of the losses, accounting for 85 per cent of the decline. Nvidia Corp co-founder Jensen Huang’s fortune fell by $20.1 billion, a 20 per cent drop, while Oracle Corp co-founder Larry Ellison saw a larger absolute loss of $22.6 billion, representing 12 per cent of his net worth. Michael Dell of Dell Inc lost $13 billion, while Binance Holdings Ltd co-founder Changpeng “CZ” Zhao saw a $12.1 billion decrease.

DeepSeek’s appeal lies in its claimed cost efficiency and performance metrics. The company reported that training its V3 model required less than $6 million in computing power using Nvidia’s H800 chips, although critics suggest total development costs could exceed $1 billion. The pricing structure particularly stands out, with DeepSeek’s R1 charging $0.55 per million input tokens compared to OpenAI’s rate of $15 per million input tokens, making it potentially 20 to 50 times more affordable.

Performance indicators have also drawn attention. According to Dev.to, a software developer community, DeepSeek achieved a 92 per cent success rate in complex problem-solving tasks, compared to GPT-4’s 78 per cent. The model has demonstrated superior logical reasoning capabilities, outperforming ChatGPT and Claude AI by seven to 14 per cent.

The development has caught the attention of industry leaders. OpenAI CEO Sam Altman acknowledged the R1 model as “very impressive,” while Google CEO Sundar Pichai emphasised the importance of taking Chinese AI development seriously. President Donald Trump characterised it as a “wake-up call” for US industries.

A distinctive feature of DeepSeek is its open-source nature. The R1 code is available to the public under the MIT License, allowing users to download, modify, and integrate it into existing systems without significant restrictions. This approach contrasts with the proprietary models developed by major US tech companies.

The market reaction highlights the concentration of investor capital in high-valuation technology stocks. Before Monday’s decline, Nvidia’s shares traded at nearly 60 times earnings, compared to the S&P 500’s average of 22. The AI boom had previously driven approximately $10 trillion in market value growth for the “Magnificent Seven” companies since ChatGPT’s launch in November 2022.

However, retail investors viewed the selloff as an opportunity. Vanda Research reported record retail purchases of Nvidia shares, totalling $562.2 million on Monday, with buy orders outnumbering sell orders by a 2:1 ratio according to JP Morgan data.

The market impact extended globally, affecting technology shares from Tokyo to Amsterdam. Dutch semiconductor company ASML and German companies Infineon and SAP experienced declines. In Japan, Nvidia supplier Advantest saw a 19 per cent drop over two days.

The development has implications for upcoming tech earnings reports from companies like Microsoft, Tesla, and Meta, with investors likely to scrutinize planned AI investments more closely. These companies have committed substantial resources to AI development, with analysts estimating US cloud companies could spend $250 billion on AI infrastructure in 2024 alone.

Monday’s market shift highlights the volatility of tech and AI investments as companies navigate increasing competition and shifting market dynamics.

This shift undermined investor confidence in the Western AI supply chain, which has been a significant driver of market growth over the past two years.

Over the same period, AI-focused companies like Meta Platforms Inc, Alphabet Inc, and Microsoft Corp have seen soaring valuations, creating substantial wealth for their founders. Since OpenAI’s launch of ChatGPT in November 2022, these companies have invested heavily in AI development, relying on advanced semiconductors and substantial energy resources to support their systems. Despite limited revenue returns, markets have rewarded these firms with record-high valuations.

Meta CEO Mark Zuckerberg recently announced plans to allocate $60 billion to $65 billion on AI projects in 2024, exceeding Wall Street projections.

While Huang and Ellison experienced significant losses, other tech billionaires remained relatively unaffected. Zuckerberg’s net worth increased by $4.3 billion as Meta’s stock rebounded, while Amazon founder Jeff Bezos added $632 million to his wealth. Huang’s net worth, however, remains notable, having increased nearly eight-fold to $121 billion since early 2023 before Monday’s losses. Zuckerberg’s wealth has surged 385 per cent to $229 billion in the same period, while Bezos’ fortune grew 133 per cent to $254 billion.

Nvidia has responded to the situation by downplaying the competitive threat, noting that DeepSeek’s advancement complies with US technology export controls. The company emphasised that inference—the process of running AI models—still requires significant numbers of Nvidia GPUs and high-performance networking.

The market reaction occurred amid broader economic developments. The Federal Reserve began its two-day policy meeting, with expectations of maintaining steady rates. Additionally, new US Treasury Secretary Scott Bessent is advocating for a gradual increase in universal tariffs, potentially starting at 2.5 per cent and rising to 20 per cent, with Trump indicating support for even higher targeted duties on products including semiconductors.

The emergence of DeepSeek has altered perceptions about China’s position in the global AI race. Experts suggest that China has reduced its technology gap with US AI models from 18 months to approximately six months.with agency inputs

Next Story
Share it