DeepSeek: Accelerating the path towards AI inevitability

Unless you have been living in an underground bunker without internet access, you have probably heard of DeepSeek over the last few weeks. As one of those rare tech developments that has managed to capture mainstream attention, the surge in interest in this objectively complex topic has led to a wide range of narratives. For brands and advertisers, separating signal from noise and understanding the implications can be challenging.

Let us start with the most important takeaway: DeepSeek is unlikely to have much (if any) direct impact on the marketing industry as it does not unlock any AI-driven use cases beyond what is already possible with existing AI models. That said, its broader impact on accelerating AI development and adoption will lead to meaningful benefits across all sectors, including marketing.

For the purposes of this article, we will focus on areas that will most impact marketers.

What is DeepSeek? 

DeepSeek is a Chinese AI company owned and funded by a hedge fund named High-Flyer. With approximately 150 employees and growing rapidly, it is not a “side project” as described by some, but a serious AI lab that has been on the radar of analysts ever since the release of the V2 model in May 2024. The company made headlines near the end of January with its free chatbot app powered by two large-language models (LLMs): DeepSeek V3, its general-purpose foundational model, and DeepSeek R1, its specialised reasoning model built from V3.

At a technical level, DeepSeek is significant as it achieved model performance comparable to that of leading Western AI labs (for instance, OpenAI, Anthropic) despite sanctions from the United States limiting access to the most advanced Nvidia chips. In AI parlance, there are two terms to consider when it comes to chip capabilities:

• Compute: The raw processing power required to train models and perform operations
• Memory: The ability to hold and organise data while working on a task

So far, models have improved largely through increasing compute, memory, data, and model size. Due to the sanctions, DeepSeek had to work with memory-constrained chips. This forced efficiency led to the development of breakthroughs in how AI models can be optimised, representing an alternate path from the "bigger is better" paradigm of simply throwing more resources at a problem.

These breakthroughs have led to definite cost efficiencies. While this is a big deal, the headlines that claim DeepSeek “does for $6 million what it took OpenAI $1 billion to do” are misleading. LLM costs can largely be split into two categories:

• Training: The costs associated with developing and training the model. This includes everything from infrastructure and compute/memory (e.g., GPUs) to R&D and actual model training runs.
• Inference: The costs of a model taking an input (e.g., a prompt) and outputting a response. This can involve consumers/businesses using the completed model or AI labs using existing models to output data to train new models.

The reported $6 million ($5.576 million, to be exact) refers to only the final training run for the V3 model (which DeepSeek states in its technical paper) and excludes all other costs incurred to reach that point. So no, V3 or R1 were not developed from scratch for $6 million.

On the other hand, DeepSeek’s cost efficiencies for inference (because of the aforementioned innovations) should be getting much more attention. Based on API pricing, inference costs for DeepSeek models are orders of magnitude cheaper than those of comparable Western competitors. Decreasing inference costs is a significantly more important enabler of AI permeation across applications and industries (including marketing/advertising).

Another key differentiator is DeepSeek's approach to open weights. While R1 may not be the absolute best model available, it is considered the leading "open weights" model. Technically speaking, its models aren’t open source. Instead, DeepSeek shares its models’ “weights”, but not everything that went into arriving at those weights (training data, process, etc). The company has gone a step further and released its models with MIT permissive licensing and detailed technical documentation, making it easier for businesses and developers to experiment with and integrate their technology.

What does this mean for brands and advertisers?

As the marginal cost of inference continues to decrease, these businesses' economic viability and scalability will skyrocket. This, in turn, will dramatically increase AI-driven innovation and utility for the marketing industry. Looking at the past for context, this is similar to how decreasing marginal costs for CPUs, internet and cloud computing enabled modern digital marketing.

The open weights and permissive licensing approach serve as a force multiplier for AI progress by lowering barriers to entry and enabling more customised solutions. This democratisation benefits startups and SMBs while opening possibilities for domain-specific adaptations, including marketing applications. It also addresses privacy concerns, especially relevant because the DeepSeek app collects more personal data than Western AI labs (please refer to the table for summary).

Many are uncomfortable with DeepSeek’s data privacy policies, particularly the idea of sending data to servers in China, where it will be subject to local laws. The good news is that while DeepSeek’s app is the most direct way to access its models, it’s not the only way. Users can run these models locally or access them through apps/platforms (Perplexity, HuggingFace, Azure) with servers outside of China, preventing unwanted data sharing and PRC-related censorship.

DeepSeek's emergence has intensified competition in AI, with China demonstrating its ability to compete with US leadership in the field. This has already prompted responses from OpenAI and Meta, who are accelerating the development of their next models. More importantly for businesses, DeepSeek's cost efficiencies have triggered pricing responses from Western labs, which will further catalyse development at the application layer where brands and advertisers stand to benefit most.

As the AI landscape continues to evolve, the indirect benefits of DeepSeek's innovations (lower costs, increased competition, and broader access to AI capabilities) will likely have a more lasting impact on marketing and advertising than any direct applications of their specific models.

---

Vincent Niou is the founder of digital marketing agency Skeleton Key.