Strategic Intelligence: Deep Dive into DeepSeek

Strategic Intelligence: Deep Dive into DeepSeek

Summary

In January 2025, Chinese AI company DeepSeek released R1, a logical reasoning large language model (LLM). Reasoning LLMs are capable of planning, coding, and mathematical problem-solving. On common AI tests in mathematics and coding, DeepSeek-V3 matched the scores of OpenAI’s GPT4o, Anthropic’s Claude 3.5 Sonnet, and Meta’s Llama 3. R1 matched OpenAI’s o1 model. Reasoning LLMs may bring us one step closer to artificial general intelligence (AGI).

Key Highlights

• The distillation of large language models (LLMs) into small language models (SLMs) could lead to thousands or tens of thousands of SLMs equipped with reasoning functionality. This will create hardware solutions that are more cost-effective, use less power, and are programmed to suit different design targets. Cloud companies will find new growth from hosting large numbers of SLMs. Smartphone makers will also benefit from SLMs.
• The US could respond to China’s AI challenge with increased research funding and subsidies. AI leadership will remain a central geopolitical issue, shaping policy and investment decisions.
• History shows that cost advantages often lead to more investment, not less. The broader AI market could benefit as more affordable AI solutions increase demand for chips and related services. The ability to pack SLMs into smartphones could further expand the market.

• This report takes a detailed look at what DeepSeek has produced and its implications for the global AI market. It also examines the current state of LLM development, including analysis of tools from Google, Meta, and OpenAI.

• DeepSeek has demonstrated the power of reinforcement learning (RL) to create a reasoning LLM. OpenAI has a leading reasoning model called o1, but the difference is in how DeepSeek accomplished the task. Reasoning LLMs will make more and more decisions on their own and will be able to improve their answers. This report will help you understand what these developments mean for the global AI industry, including analysis of the potential impact on leading vendors.