Recently, DeepSeek has made waves in the artificial intelligence (AI) community by introducing a more efficient way to scale large language models (LLMs), challenging the dominance of OpenAI and other closed-model vendors. This development has sparked a heated debate on the merits of open versus closed AI approaches, but amidst the noise, a more critical issue has emerged: the importance of customer trust in the emerging AI market.

DeepSeek's achievement is significant because it demonstrates that AI entry costs need not be prohibitively high. By optimizing compute and memory usage, DeepSeek's R1 model can process one million output tokens for just $2.19, compared to OpenAI's o1 model, which costs $60. This price difference has far-reaching implications for the democratization of AI, making it more accessible to a broader range of developers and users.

However, the real significance of DeepSeek's innovation lies in its potential to disrupt the existing AI landscape. OpenAI's closed approach has raised concerns about customer trust, with the company claiming ownership of the output generated by its models. This "techno-feudalism on steroids," as Me & Qi cofounder Arnaud Bertrand puts it, has serious implications for the adoption of AI in the enterprise sector.

Enterprises are hesitant to adopt AI at scale because they don't trust LLM vendors with their data. OpenAI's stance on model output ownership has exacerbated this mistrust, making it clear that the company prioritizes its own interests over those of its customers. In contrast, AWS and Microsoft have built successful cloud businesses by establishing trust with their enterprise customers, demonstrating that trust is a critical factor in the AI market.

As the AI market continues to evolve, it's clear that the winners will be those vendors that prioritize customer trust. Open source approaches can help, but ultimately, it's the vendor's attitude towards customer data that will determine their success. The AI companies that focus on establishing trust with their customers will be better positioned to capitalize on the emerging market opportunities.

In the rush to dominate the AI market, companies like OpenAI have forgotten that the real, biggest market has yet to emerge. Tim O'Reilly's warning to AI company leaders and their investors to avoid prioritizing monopoly power over customer needs is particularly relevant in this context. By focusing on customer trust, AI vendors can create a more sustainable and equitable market that benefits everyone involved.

In conclusion, DeepSeek's innovative approach to LLMs has highlighted the importance of customer trust in the AI market. As the market continues to evolve, it's clear that the winners will be those vendors that prioritize trust, transparency, and customer control. The AI companies that adapt to this new reality will be better positioned to succeed in the long term.