As many businesses are now racing to apply AI to their business because they can simplify operations, improve automation, and develop innovative products. Your choice of AI models will determine efficiency, scalability, and integration with existing systems. While the market has been dominated by closed-source models like OpenAI GPT. However, open-source models like Deepseek have now come onto the scene with a bang.
The day Deepseek launched it caused one of the biggest stock market crashes in the history of the US stock market. With over $593 billion being lost in Nvidia's valuation alone. DeepSeek, a Chinese startup founded in 2023, offers its AI models as open source, including its R1 reasoning model, allowing for free use and adaptation. The technology industry took notice of DeepSeek for several reasons, but its development cost of under $6 million and cost-efficient hardware stood out.
What is Deepseek?
DeepSeek is an AI model (a chatbot) that functions similarly to ChatGPT, which enables users to perform tasks like coding, reasoning, and mathematical problem-solving. It is powered by the R1 model, which boasts 670 billion parameters. Now, making it one of the largest open-source language models. Currently, it has two models, which are the V3 and R1.
Deepseek R1 model excels in reasoning by producing responses more carefully and incrementally. This is to mimic the human thought process. This way of response creation reduces memory usage. Thus making it more cost-effective than other models in the market. Hence, makes it stand out among other chatbots. The figure quoted to make this model is just $6 million, which is a fraction of the $100 million-plus price tag of ChatGPT.
DeepSeek R1 has made its code open-source, though it still keeps the training data proprietary. This transparency allows for verification of the company’s claims. Moreover, the model’s computational efficiency promises faster and more affordable AI research, opening doors for broader exploration. This accessibility may also facilitate deeper investigations into the mechanics of large language models (LLMS).
Deepseek Maker
DeepSeek was founded in December 2023 by Liang Wenfeng, who launched the first large language model the following year. Liang, an alumnus of Zhejiang University with degrees in electronic information engineering and computer science, has emerged as a key figure in the AI industry worldwide due to his continued advancement in the field and industry.
Different from other AI founders. Liang himself has a huge background in finance. He is the CEO of High-Flyer, a hedge fund specialising in quantitative trading, which leverages AI to analyse financial data and make investment decisions. In 2019, High-Flyer became China’s first quant hedge fund to raise over 100 billion yuan (£10 million).
Now called Sam Altman of China, Liang has been vocal about China’s need to innovate rather than imitate AI. In 2019, he emphasised the need for China to advance its quantitative trading sector to rival the US. He believed that the true challenge for Chinese AI was transitioning from imitation to innovation, a shift that required original thinking.
Key innovations of the DeepSeek-V2 model
DeepSeek-V2 introduces several key architectural advancements. It employs a novel MoE architecture and MLA attention mechanism. So to explain further, an MoE is a Mixture-of-experts (MoE) architecture. Which activates only a subset of the model’s parameters and concurrently minimises the computational resources required for processing the query. In simple terms, it means that instead of having a single, massive neural network, the model consists of multiple smaller “expert” networks, each specialising in different input aspects. During processing, only a subset of these experts is activated for each input, making the computation more efficient.
The other innovation is the MLA attention mechanism. It is a novel attention mechanism that significantly reduces the memory footprint of the model. Traditional attention mechanisms require storing large amounts of information, which can be computationally expensive. MLA compresses this information into a smaller “latent” representation, allowing the model to process information more efficiently.
Training innovations in DeepSeek
DeepSeek uses a different approach to train its R1 models than what is used by OpenAI. The training involved less time, fewer AI accelerators and less cost to develop. DeepSeek's aim is to achieve artificial general intelligence, and the company's advancements in reasoning capabilities represent significant progress in AI development. In this section, we are going to outline the key innovations in terms of training.
- Reinforcement learning: DeepSeek used a large-scale reinforcement learning approach focused on reasoning tasks.
- Reward engineering: Researchers developed a rule-based reward system for the model that outperforms neural reward models that are more commonly used.
- Distillation: Using efficient knowledge transfer techniques.
- Emergent behaviour network: DeepSeek's emergent behaviour innovation is the discovery that complex reasoning patterns can develop naturally through reinforcement learning without explicitly programming them.
The sudden rise of Deepseek
The significance of DeepSeek lies in its potential to dramatically transform AI’s tech and financial landscape. When tech leaders in the US were busy investing in nuclear energy to keep their power-guzzling data centres running, DeepSeek achieved the same objectives without the fuss. AI development itself consumes immense resources, exemplified by Meta’s $65-billion investment in developing technology.
With the advent of Deepseek, it shows how at par AI capabilities can be achieved with so much lower cost and less sophisticated hardware. This notion basically destroys the bubble that the development of AI models needs a huge amount of investment. The availability of AI models at a fraction of the cost and with less sophisticated chips can increase their usage by industries manifold, enhance productivity, and foster unprecedented innovation.
DeepSeek cyberattack
DeepSeek's popularity has not gone unnoticed by cyberattackers. On Jan. 27, 2025, they reported large-scale malicious attacks on its services, forcing the company to temporarily limit new user registrations. The timing of the attack coincided with their AI assistant app overtaking ChatGPT as the top downloaded app on the Apple App Store.
Despite the attack, DeepSeek maintained service for existing users. The issue extended into Jan. 28, when the company reported it had identified the issue and deployed a fix. They have not specified the exact nature of the attack, though widespread speculation from public reports indicated it was some form of DDoS attack targeting its API and web chat platform.
Conclusion
DeepSeek’s rise disrupts AI by proving advanced models can be built affordably ($6M) with open-source frameworks, challenging costly giants like OpenAI. Its MoE architecture and MLA attention boost efficiency, democratizing AI access. Despite a 2025 cyberattack, its resilience and founder Liang Wenfeng’s vision highlight China’s shift from imitation to innovation, reshaping global AI dynamics.
If you decide to enrol in a course at the College of Contract Management, you’ll boost your skills in CAD design and project management. This education is a great way to jumpstart your journey. Plus, it sets you up to become a leader in the industry. You'll learn to manage projects and create innovative designs. With all these skills, you can truly contribute to the ongoing evolution of the construction sector. So, why not take the first step towards a new set of skills today? Take your construction career to the next level with CAD design and AI.
