‍

The release of DeepSeek’s latest AI models, DeepSeek-V3 and DeepSeek-R1, has triggered a sharp sell-off in AI-related stocks, reflecting market concerns over the shifting competitive landscape. These models are reported to match the capabilities of ChatGPT-4o while being significantly cheaper to train and requiring less computing power. While the initial market reaction has been negative, we believe this development is ultimately a positive for the AI sector, driving greater adoption, efficiency, and long-term growth.

‍

The Impact of DeepSeek’s AI Models

‍

Lowering Costs and Expanding Accessibility

One of the biggest barriers to AI adoption has been the high cost of training and inference. DeepSeek has demonstrated that state-of-the-art models can be developed at a fraction of the cost, with significantly improved efficiency. The open-source nature of these models will likely accelerate innovation, allowing smaller players, research labs, and start-ups to develop AI applications without requiring the vast financial resources previously needed.

From an investment perspective, this shift broadens the AI ecosystem beyond the dominance of US Big Tech, which has so far been the primary driver of AI advancements due to its ability to invest heavily in Graphics processing units (GPU) infrastructure. While this may seem like a threat to incumbents, history suggests that lower costs and increased accessibility ultimately expand markets rather than shrink them.

‍

Addressing Energy Supply Constraints

Another major concern around AI expansion has been energy consumption. AI training and inference require vast amounts of computing power, leading to concerns about whether existing energy infrastructure can support long-term growth. DeepSeek’s more energy-efficient models help alleviate this concern by demonstrating that GPU usage can be further optimised.

This suggests that the energy bottleneck may not be as severe as previously feared, allowing AI to scale without immediate infrastructure constraints. While energy consumption will remain an important factor, improved model efficiency could delay the need for drastic energy infrastructure investments and make AI adoption more sustainable.

‍

Investment and Industry Implications

Given the reduced cost and energy barriers, we expect the AI industry to accelerate in several key ways:

‍

1. Larger, More Ambitious Models

• AI model development is often constrained by training costs and time. With these constraints easing, we could see a shift towards even larger, more powerful AI models, expanding capabilities and use cases.

‍

2. Broader Adoption and Increased AI Use Cases

• Lower costs will make AI more viable across a wider range of industries, from healthcare and finance to smaller-scale applications by start-ups and individual researchers.

‍

3. Increased Model Iteration and Retraining

• The ability to retrain models more efficiently will lead to faster improvements in AI, as developers can afford to iterate and refine their models more frequently.

‍

4. Lower Barriers to Entry, Increased Competition

• As state-of-the-art AI becomes more accessible, new competitors will emerge, driving innovation and diversification in the market. While this may challenge the dominance of established tech giants, it could also lead to new investment opportunities in emerging AI players.

‍

Despite initial concerns, US Big Tech is unlikely to scale back AI investment. While some may temporarily slow spending to assess the competitive landscape, ultimately, they will continue investing heavily in GPUs, AI infrastructure, and new AI use cases. Choosing to limit investment now could risk falling behind competitors, something major tech firms will be keen to avoid.

‍

Market Reaction and Investment Outlook

The sharp sell-off following DeepSeek’s announcement suggests that the market has reacted with short-term pessimism, focusing on increased competition rather than long-term industry expansion. This is a classic example of the market opting to “shoot first and ask questions later.”

At Blue Whale, we believe this presents an opportunity rather than a threat. While volatility is expected, the broader trends remain intact: AI adoption is set to accelerate, demand for semiconductors and AI infrastructure will grow, and efficiency gains will drive new applications and markets.

We remain confident in our long-term investment thesis on AI infrastructure and will be looking to take advantage of this market dislocation by investing in our preferred AI holdings at more attractive valuations.

‍

Conclusion

DeepSeek’s advancements mark an important shift in the AI industry, lowering barriers to adoption, increasing efficiency, and enabling broader competition. While market reactions have been negative in the short term, we see this development as a positive catalyst for the sector. The AI industry is still in its early stages, and developments like these only reinforce its long-term growth potential.

We will continue to monitor the situation closely and adjust our strategy accordingly. If you would like to discuss this further, please feel free to reach out.

‍

Comments from the investment team at Blue Whale, managers of the Blue Whale Growth Fund which features on the EXE Capital Management buy list.

Original wording summarised using AI.

The views are those of the author only. The above does not constitute a recommendation to buy the fund and advice should be sought from your financial advisor as to the appropriateness of this fund in your portfolio. The value of investments can fall as well as rise. Past performance is no guarantee of future returns.

‍