
On Monday, January 27, 2025, the tech world witnessed a seismic shift as news of a Chinese AI company called DeepSeek sent shockwaves through global financial markets. This unexpected development led to a staggering $1 trillion wipeout in the U.S. stock market, leaving investors and industry experts scrambling to understand the implications. DeepSeek saw an incredible surge in user signups and faced a cyber attack.
The DeepSeek Phenomenon
DeepSeek, a relatively unknown Chinese AI startup, burst onto the scene with its latest reasoning model, DeepSeek R1. This new AI model quickly gained attention for its impressive capabilities, rivaling or even surpassing those of industry leaders like OpenAI’s ChatGPT.
What sets DeepSeek apart?
- Free and Open-Source: DeepSeek made its model available for free as both a web and mobile application.
- Customizable or make your own AI model: BIG WIN!Users can operate, download and modify the model locally, addressing privacy concerns.
- Cost-Efficient: EVEN BIGGER WIN! Reports suggest that training the DeepSeek V3 model cost only $5.58 million, a fraction of what major U.S. AI firms reportedly spend.
- Innovative architecture: DeepSeek employs a “Mixture-of-Experts” (MoE) architecture, which activates only a small portion of the model’s parameters at any given time, leading to efficient computing power usage. Imagine a company with a team of specialists, each expert in a different area. When a complex project comes in, instead of everyone working on everything, a smart manager (let’s call them the “router”) decides which experts are best suited for each part of the project. The beauty of this system is that it’s efficient. Not every expert needs to work on every task, saving time and resources.
- Phrase Interpretation: Instead of relying solely on individual keywords, DeepSeek can understand and process entire phrases or sentences to grasp the full context of a search.
DeepSeek Vs ChatGPT at a glance
| Feature | DeepSeek | ChatGPT |
|---|---|---|
| Cost | $0.50/month for premium; $0.55 per million input tokens | $20/month for premium; $15 per million input tokens |
| Open Source | Yes | No |
| Development Cost | Approximately $5.58 million training cost | Over $100 million |
| Strengths | Technical tasks, coding, logical reasoning, math problems | Creative writing, conversational tasks, general knowledge |
| Customization | Can be run and modified locally | Offers customized GPTs |
| API Pricing | $0.55 per million input tokens, $2.19 per million output tokens | $15 per million input tokens, $60 per million output tokens |
| Hardware Requirements | Minimal (can run on older Nvidia GPUs) | High-end GPUs |
| Reputation | Newer, gaining traction | Established, widely trusted |
| Contextual Understanding | Strong but slightly less accurate | Excellent |
| Content style | Real time searches & summaries | Creative |
Market Plunge
- Nasdaq Plunge: The tech-heavy Nasdaq index fell nearly 600 points, or almost 3%, by late morning on January 27, 2025.
- Nvidia’s Freefall: Nvidia, a leading AI chip manufacturer, saw its stock plummet by over 13%, wiping out nearly $600 billion in market value.
- Widespread Tech Selloff: Other tech giants and AI-related stocks, including Broadcom, Arm, Microsoft, and Alphabet, also experienced significant losses.
Investors of Energy, AI and Electronics(AI chips) companies are revisiting their investment strategy after this big shock wave.
Expert Opinions
Industry leaders and experts weighed in on DeepSeek’s impact:
- Marc Andreessen, investor and engineer, called DeepSeek R1 “AI’s Sputnik moment,” drawing parallels to the space race.
- Satya Nadella, Microsoft’s CEO, acknowledged the impressive nature of DeepSeek’s model, particularly its computational efficiency.
- Dan Ives, Wedbush Securities analyst, highlighted the stark contrast in development costs between DeepSeek and its U.S. counterparts.
Speculation
Are we safe with OpenSource Cheaper AI models?
DeepSeek being Open Source poses major concerns : Malicious usage, Ethical concerns, Unintended Consequences. Anyone can download it and run it locally.
Should US tech companies revisit their approach to reduce cost?
Companies have declared Millions of dollars of investment in AI. By 2025, global AI investment is projected to reach $200 billion, with the US accounting for a substantial portion of this amount.DeepSeek’s ability to create a high-performance AI model at a fraction of the expected cost challenged assumptions about the necessary investment in AI development.
Are we overreacting to the hype?
Maybe, maybe not! Each AI model has its pros and cons. There are questions on how a feat like this was accomplished by Deepseek at such a low cost. Is this real?
DeepSeek’s Cost Claims maybeToo Good to Be True?
Many analysts argue that DeepSeek’s reported figure doesn’t tell the whole story. The $5.6 million likely only covers the final training run, omitting crucial expenses for R&D, Data acquisition, OpEx, Infrastructure, hardware costs(potentially in 100s of millions).
There are questions about data quality and potential reliance on existing models and suspicions that DeepSeek-V3 may have been trained using ChatGPT outputs.
Political and Ethical Concerns such as Data privacy issues similar to those raised about ByteDance and TikTok, Potential for biased outputs, especially on topics sensitive to China, Lack of ethical guardrails are all potential concerns.
While DeepSeek’s innovations are undoubtedly impressive, the AI community is right to approach these cost claims with a critical eye. As the dust settles, a more complete picture of DeepSeek’s true development costs and methods may emerge, providing valuable insights for the future of AI development.
The Aftermath and Recovery
While the initial market reaction was severe, some recovery is observed. By January 28, 2025, several affected stocks showed signs of rebounding:
- Nvidia shares rose about 5% in premarket trading.
- Broadcom and TSMC saw increases of 3.4% and 3.2%, respectively.
And guess what, Google trend has already picked up on Super Bowl. People have moved on.

However, the long-term implications of DeepSeek’s emergence remain to be seen, as investors and companies reassess their strategies in light of this new player in the AI landscape.As the dust settles, it’s clear that DeepSeek has not only disrupted the stock market but also challenged the status quo in AI development. This event serves as a reminder of the rapid pace of innovation in the tech sector and the potential for unexpected players to reshape entire industries overnight.
One Comment Add yours