DeepSeek, the recently launched open-source artificial intelligence from China, has made a splash in the market following a $5.6 million investment in its development, positioning itself as a more efficient alternative to giants like OpenAI, Google, Microsoft, Amazon, Apple, and Meta. Its impact is already being felt in the industry, challenging the subscription-based business model that major tech companies have implemented.
According to Statista, the development of Gemini, Google’s AI, has required an investment of between $30 million and $190 million, not including salaries, while ChatGPT-4 has exceeded $100 million in costs, according to statements from Sam Altman. In light of these figures, the Chinese model emerges as a high-quality option at a fraction of the cost, generating a strong impact in the sector.
Enric Quintero, CEO of Datarmony, describes this scenario as “a business earthquake,” as the offering of free and open-source AI by DeepSeek challenges the traditional subscription-based business model, which could alter the commercial strategies of major tech firms. Additionally, the openness of its code allows other players to develop more advanced models based on its technology, potentially accelerating the evolution of the sector significantly.
Meanwhile, Felipe Maggi, CDO of Datarmony, emphasizes that this advance “greatly reduces the cost of training AI models and weakens the barriers to entry for institutions with expert talent but less economic capability.” This means that research centers, startups, and other organizations will be able to develop competitive models without the need for massive investments in hardware and computational resources.
In terms of privacy and regulatory compliance in the European Union, DeepSeek faces the same challenges as ChatGPT and Gemini. Data processed on its servers should not be used to manage personal or confidential data. However, Quintero and Maggi clarify that if it is used in private environments or on servers located within the EU, it would avoid the transfer of data to China, alleviating one of the main concerns surrounding the use of AI developed in that country.
Another factor that makes DeepSeek an attractive alternative is its energy efficiency and accessibility. Unlike its Western counterparts, this model does not require specialized or expensive hardware, being able to operate on standard equipment. This represents a significant advantage in terms of sustainability and democratizing access to artificial intelligence.
The impact of DeepSeek has already had repercussions in the financial market. Companies like Google, Microsoft, Apple, Amazon, and OpenAI have invested millions in hardware for the development of their AI models. However, the emergence of a more accessible alternative that is less dependent on costly infrastructures has raised doubts about the profitability of these investments. In this regard, the market has reacted decisively: NVIDIA, the leading chip supplier for training AI models, has lost $600 billion in market value in recent days.
With a free, open, and efficient model, DeepSeek is challenging the status quo of artificial intelligence, forcing tech giants to rethink their strategies in a rapidly evolving sector. Its arrival not only signifies a shift in market competitiveness but also opens new opportunities for the development and accessibility of AI across various sectors.