In 2017, a seminal paper by a team of eight Google researchers titled “Attention Is All You Need,” commonly known as the Transformer Paper, laid the foundation for a technological revolution. This paper introduced the concept of self-attention mechanisms in Transformers, a critical advancement that allowed computers to understand long-range dependencies and the semantic structure of human languages.
Since then, technologists have used transformer architecture as a backbone to develop Large Language Models (“LLMs”), capable of interpreting natural language and even generating text, images, and videos. These LLMs mark the advent of what the industry now recognizes as generative AI.
The investment landscape in generative AI is both robust and complex, with considerable capital being allocated to startups in the sector. However, the understanding of where the value will accrue in this rapidly evolving field remains incomplete by many. In this landscape of uncertainty, our current insight leads us to believe that LLMs represent a promising area of focus and will likely form the foundational layers powering intelligent end-user experiences. These specialized models, with their capacity for contextual understanding and application across diverse industries, hold a key to unlocking significant value within the broader ecosystem of generative AI.

Generative AI Market Size, 2023-2032 ($B)

At this stage of the technology curve, the most exciting use case for generative AI sits within the enterprise segment, where the innovative technology can be used to supercharge existing processes and increase worker efficiencies by up to 40%.
Cohere, a leading player in the enterprise LLM space, has identified and seized the opportunity within this burgeoning field. It has created AI foundational models that have the power to redefine enterprise software and workflows through the use of AI by focusing on customer flexibility, data privacy, and industry-leading accuracy.

Enterprise Leaders’ View on Most Disruptive Technologies Today

In a world where generative AI is transforming rapidly, Cohere’s commitment to a more deliberate approach has enabled it to engineer solutions that not only align with enterprise needs but also maintain neutrality across different cloud environments. With 91% mean accuracy and the second highest HELM ranking, Cohere stands out as a category leader. Also, given how resource intensive it can be to train foundational models, Cohere’s current leadership and success should further solidify the company’s moat, leading to high barriers of entry and competitive defense.

The success of Cohere isn’t an accident but the result of meticulous planning, deep understanding, and the right team. Aidan Gomez, one of the original authors of the Transformer Paper, leads Cohere. His expertise, combined with a seasoned team proficient in technological innovation, gives Cohere an edge in navigating the complex landscape of generative AI.

Cohere’s focus on customer flexibility and data privacy reflects a mature understanding of the market’s needs, a vision aligned with the challenges and expectations of enterprise clients. It’s not just about creating groundbreaking technology; it’s about building a business that resonates with the values and demands of the contemporary world.

At Mirae Asset, we recognize the complex and dynamic nature of the generative AI landscape, and we’re thrilled to partner with Aidan and his team. Cohere’s focus on technological innovation, ethical alignment, and a practical understanding of market dynamics makes it a vital player in this emerging field.

To learn more please contact us at