Found: Getting realistic about AI’s potential with Nick Frosst from Cohere

Business Integration

Enterprise AI is rapidly evolving, and companies like Cohere are at the forefront, helping businesses integrate AI without the need to build and store their own models. Nick Frosst, co-founder of Cohere, explains that the focus has shifted from explaining what a language model is to demonstrating how their models can meet specific business needs, including data security and multilingual capabilities 1. This shift highlights the growing acceptance and understanding of AI's practical applications in the business world.

Now the conversations are how is your large language model good for us in particular? And we can talk about our focus on data security, on privacy. We can talk about our multilingual capabilities, our models really good at more than English.

--- Nick Frosst

Cohere's commitment to making AI useful for enterprises is evident in their partnerships, such as with Fujitsu, to develop models that cater to specific linguistic needs, demonstrating their practical approach to AI integration 2.

   

Realistic Expectations

Balancing the hype around AI with realistic expectations is crucial. Nick emphasizes that while AI has immense potential, it is not a panacea for all problems. He uses AI in his daily workflow for tasks like summarizing documents and solving research questions, showcasing its practical benefits 3. However, he cautions against extreme rhetoric on both sides, advocating for a balanced view of AI's capabilities and limitations.

I think it can add a lot of value, but I don't think it's going to bring about the death of all humans.

--- Nick Frosst

This realistic perspective is refreshing in an industry often characterized by exaggerated claims, and it underscores the importance of focusing on tangible, achievable outcomes 4.

   

Challenges and Solutions

Implementing AI tools in businesses comes with its own set of challenges. Nick points out that the availability of language data on the web varies, making it harder to develop models for less-represented languages 5. Despite these challenges, Cohere has successfully deployed models that automate tedious tasks, such as extracting data from financial PDFs, significantly improving efficiency for their clients.

We had to make the model really good at those boring things. We had to figure out the right deployment options, had to do all kinds of things that don't really have a lot to do with AGI or with neural nets even.

--- Nick Frosst

Their focus on practical solutions and customer needs has also resonated well with investors, enabling them to secure significant funding to continue their mission 6.