Topic for Feb. 6. ( Thur )

[Alex !]

Deep Seek

DeepSeek is a Chinese artificial intelligence company that develops open-source large language models (LLMs). Based in Hangzhou, Zhejiang, it is owned and funded by Chinese hedge fund High-Flyer, whose co-founder, Liang Wenfeng, established the company in 2023 and serves as its CEO.

The DeepSeek-R1 model provides responses comparable to other contemporary large language models, such as OpenAI's GPT-4oand o1. It is trained at a significantly lower cost—stated at US$6 million compared to $100 million for OpenAI's GPT-4 in 2023.

Tumbling stock market values and wild claims have accompanied the release of a new AI chatbot by a small Chinese company. What makes it so different?

The release of China's new DeepSeek AI-powered chatbot app has rocked the technology industry. It quickly overtook OpenAI's ChatGPT as the most-downloaded free iOS app in the US, and caused chip-making company Nvidia to lose almost $600bn (£483bn) of its market value in one day – a new US stock market record.

It is likely that, working within these constraints, DeepSeek has been forced to find innovative ways to make the most effective use of the resources it has at its disposal.

Reducing the computational cost of training and running models may also address concerns about the environmental impacts of AI.

The data centres they run on have huge electricity and water demands, largely to keep the servers from overheating. So, increasing the efficiency of AI models would be a positive direction for the industry from an environmental point of view.

Of course, whether DeepSeek's models do deliver real-world savings in energy remains to be seen, and it's also unclear if cheaper, more efficient AI could lead to more people using the model, and so an increase in overall energy consumption.

What has surprised many people is how quickly DeepSeek appeared on the scene with such a competitive large language model – the company was only founded by Liang Wenfeng in 2023, who is now being hailed in China as something of an "AI hero".

The latest DeepSeek model also stands out because its "weights" – the numerical parameters of the model obtained from the training process – have been openly released, along with a technical paper describing the model's development process. This enables other groups to run the model on their own equipment and adapt it to other tasks.

This relative openness also means that researchers around the world are now able to peer beneath the model's bonnet to find out what makes it tick, unlike OpenAI's o1 and o3 which are effectively black boxes. But there are still some details missing, such as the datasets and code used to train the models, so groups of researchers are now trying to piece these together.

DeepSeek is potentially demonstrating that you don't need vast resources to build sophisticated AI models. My guess is that we'll start to see highly capable AI models being developed with ever fewer resources, as companies figure out ways to make model training and operation more efficient.

Up until now, the AI landscape has been dominated by "Big Tech" companies in the US – Donald Trump has called the rise of DeepSeek "a wake-up call" for the US tech industry. But this development may not necessarily be bad news for the likes of Nvidia in the long term: as the financial and time cost of developing AI products reduces, businesses and governments will be able to adopt this technology more easily. That will in turn drive demand for new products, and the chips that power them – and so the cycle continues.

It seems likely that smaller companies such as DeepSeek will have a growing role to play in creating AI tools that have the potential to make our lives easier. It would be a mistake to underestimate that.

Deep Seek vs ChatGPT

DeepSeek has established itself as a notable challenger to the widely adopted ChatGPT, bringing a fresh perspective to AI language models. As an open-source alternative, DeepSeek has drawn significant attention for its impressive capabilities and cost-efficient approach, particularly excelling in technical and mathematical domains. While ChatGPT maintains its position with versatile features and user-friendly interface, DeepSeek's emergence gives us a compelling reason to consider it as a real option.

DeepSeek has shown impressive capabilities in technical tasks, particularly excelling in mathematics where it achieves a 90% accuracy rate - notably higher than many competitors. This makes it particularly valuable if you are working on technical problems. ChatGPT, however, demonstrates stronger capabilities in understanding context and providing more nuanced responses across a broader range of topics.

DeepSeek takes an open-source approach, meaning it's freely available and can be modified by the community. This is particularly valuable if you want to understand or customize the underlying technology. ChatGPT operates on a freemium model, offering basic features for free but requiring a subscription for advanced capabilities.

DeepSeek offers more extensive customization options. However, this comes with a steeper learning curve and requires some technical expertise. ChatGPT prioritizes user-friendliness, offering a more polished experience that's accessible even to those just starting their data science journey.

ChatGPT excels at producing engaging, conversational content with rich context - perfect for explaining complex data concepts to non-technical stakeholders. DeepSeek, on the other hand, shines in technical writing scenarios, producing precise, formal documentation that's particularly valuable for data project documentation and technical specifications.

ChatGPT offers comprehensive code assistance, providing detailed explanations alongside its code suggestions - making it an excellent learning tool for those new to data science. DeepSeek takes a more direct approach, with faster code generation and a modular style that's especially useful when you need quick, efficient solutions for specific coding challenges. Many developers have found success using DeepSeek for rapid prototyping and ChatGPT for understanding complex implementations.

ChatGPT excels at generating multiple diverse approaches to a problem, helping you explore various analytical possibilities. DeepSeek typically provides fewer but more thoroughly developed solutions, diving deep into a single approach - particularly useful when you need to flesh out a specific data strategy in detail.

ChatGPT provides comprehensive, tutorial-style explanations that work well for learning new concepts. It excels at breaking down complex topics into digestible pieces. DeepSeek focuses more on precision and conciseness, making it particularly effective for quick reference and fact-checking during your data projects. Its technical accuracy is especially valuable when researching specific methodologies or algorithms.

DeepSeek stands out for its cost-effectiveness, using energy-efficient hardware and edge deployments to keep operational costs low. Being free to use, it's an excellent resource if you are working with limited budgets. ChatGPT's subscription model, while more expensive, offers consistent performance and advanced features that can be valuable for professional data work.

Privacy and ethical concerns

This aspect is particularly important when working with sensitive data. ChatGPT follows Western data protection standards, making it a safer choice for projects requiring strict data privacy compliance. DeepSeek's data storage practices and content moderation policies might raise concerns for certain types of projects, especially those involving sensitive information or requiring unrestricted analytical discussions.

Each feature comparison reveals trade-offs that matter in different data science scenarios. The key is matching these capabilities to your specific needs and requirements.

For technical users and developers

If you're focused on coding, data analysis, and technical documentation, DeepSeek offers compelling advantages. Its open-source nature allows for customization and integration into your development workflow, while its cost-effectiveness makes it particularly attractive for individual developers or small teams. However, be prepared to implement additional verification steps for complex analyses and be aware of potential limitations in handling politically sensitive data topics.

For general users, businesses, and content creators

If you're working in a business environment or need to create content that explains data concepts to stakeholders, ChatGPT might be your better choice. Its refined language generation and user-friendly interface make it excellent for creating documentation, reports, and presentations. The consistent performance across various tasks means you can rely on it for both technical and non-technical communication.

For privacy-conscious users

For those working with sensitive data or in regulated industries, ChatGPT's adherence to Western data protection standards makes it the safer choice, as I mentioned earlier. While both tools have their privacy considerations, ChatGPT's more transparent data handling policies and stricter compliance with international privacy regulations make it more suitable for professional environments where data security is paramount.

The key is to align your choice with your primary use case: If you're focused on technical development and cost isn't a major constraint, DeepSeek's capabilities might serve you well. However, if you need a more versatile tool that can handle both technical and communication tasks while maintaining high privacy standards, ChatGPT would be the more appropriate choice.

Therefore If you're focused on specialized technical tasks and value customization options, DeepSeek's open-source framework and mathematical precision make it an excellent choice. However, if you need a well-rounded solution with strong privacy features and user-friendly interface, ChatGPT offers a more polished experience. Whichever tool you choose, both platforms represent significant advancements in AI technology, each contributing unique value to the field of data science and development.