Anders Romare, Chief Digital and Information Officer and SVP of Global Data, and IT at Novo Nordisk, speaks with Lana Feng, Co-founder and CEO of Huma.Ai, in a video interview about upscaling the workforce for generative AI (generative AI), successful collaboration, addressing hallucinations, and the need for a dialogue.
Novo Nordisk is a multinational pharmaceutical company headquartered in Bagsværd, Denmark.
As a leading pharmaceutical company it has been working with AI and ML for many years, says Romare. Further, he plans on accelerating that journey.
As part of the acceleration, Romare mentions scouting talents to upscale the workforce to be generative AI savvy. He adds that there must be a combination of scientific and market understanding with generative AI in the workforce.
Adding, Romare states that the organization focuses on balancing between recruiting new talent and upskilling the internal workforce. Moreover, he stresses the need to partner with various companies to be part of an external ecosystem of knowledge in the field of generative AI.
Having sharp focus and clear ideas about the organizational factors that add to the value chain is critical to ensure a successful collaboration, says Romare. When it comes to drug discovery or the marketing phase, it boils down to zooming in on a particular problem and providing a well-defined solution.
In continuation, Romare maintains that giants like Microsoft and Amazon have a broader impact on knowledge workers. He states that teaming up with the giants is not similar to collaborating with a startup company. Therefore, the organization has a dedicated part that is skilled in engaging with startups to drive innovation together.
Adding on, he affirms having a classical funnel mechanism where the organization can start small, take ideas, and eventually mature into something that can be scaled or aborted if it does not work.
When asked about the potential use cases for generative AI, Romare responds that it hugely depends on what phase the company is in. From the pharma perspective, he says that generative AI can be leveraged in the drug discovery phase that would benefit future patients.
However, he asserts that the organization currently has a high demand for core products, due to which there is no need to increase demand. Therefore, generative AI would not be prioritized in commercial space at the moment.
Moving forward, Romare affirms that a lot of organizational focus is drawn towards the research and development phase to ensure bringing in new innovative medicines. Secondly, he maintains that efforts must be put to support the office workers, with generative AI tooling. Romare shares that there are tangible examples where people using generative AI tooling from different vendors drive efficiency in work.
Highlighting challenges, he cautions to be mindful about applying generative AI as it can be easily trained in a biased manner. Romare reminds people of the socio-ethical responsibility that comes with it. He further refers to the AI Legislation Act in Europe, which is coming in place with clear rules, regulations, and fines in case of violation.
Commenting on hallucination in Large Language Models (LLMs), Romare notes that the organization is still in the early phases of deploying and has decided on a tiered approach. Explaining further, he says that the aim is to start with use cases that are less sensitive from an output perspective while the data quality is well understood.
By doing that, the organization is putting together groups that have an ethical and quality dimension to it, while later expanding to more advanced use cases over time. Thus, he affirms that the organization is keen to avoid anything that is ethically wrong or provides hallucinations.
In conclusion, Romare says that a healthy dialogue between authorities and industries is good as it increases understanding on both ends around usage and limitations around AI. Also, the company carefully adheres to the legislation, but there is a normal interplay of roles, he notes.
CDO Magazine appreciates Anders Romare for sharing his insights and data success stories with our global community.
