(US & Canada) VIDEO | Data Is All About Asking and Answering Questions — US Navy CDAIO

Duncan McCaskill, Chief Data and AI Officer at the U.S. Department of the Navy, speaks with Katya Mijatovic, Principal Data Scientist at Data Society and CDO Magazine Editorial Board Member, about the cultural aspect of data, training the workforce, transformational technologies, and leveraging generative AI at the department.

Mijatovic interviewed McCaskill on the sidelines of the recently conducted AFCEA CeVA Data Centric Summit.

Sharing key points from his presentation at the summit, McCaskill says that people think of data and artificial intelligence as technical things. However, he maintains that while there are some technical components they are primarily about culture and people.

McCaskill elaborates that data is about being able to answer and ask questions. He shares that the hardest part is getting the organization and the people in the frame of mind to be able to do that cooperatively.

When asked about his approach to keeping the staff updated and the culture evolving with technology, McCaskill says that organizationally, the data office sits within the chief information office and has a team of five civilians. While being able to handpick talent for the team, it also has direct hiring options for various roles.

Apart from picking talent, McCaskill says that the workforce also undergoes rigorous training, upskilling the entire department to achieve data and analytical fluency.

Speaking about a couple of technologies that have been highly transformational for the organization and in general, McCaskill highlights the significant impact of cloud technologies on cooperation and communication across Department of Defense elements.

Cloud computing emerged as a game changer, enabling rapid deployment and scalability without the burden of managing costly infrastructure.

Additionally, he mentions text analysis, particularly advancements in natural language processing (NLP) and generative AI. The combination of cloud capabilities and advanced text analysis has unlocked unprecedented possibilities in handling text-based communications, including documents and emails, which were impossible a decade ago

Responding to Mijatovic’s question about a technology he is most looking forward to in the next decade, McCaskill reiterates generative AI and capabilities like Retrieval-Augmented Generation (RAG) with the ability to apply foundational models securely to different mission sets.

He cites the example of using various technical manuals for repairing an airframe. McCaskill says that if executed correctly while preventing hallucinations, the model can act like an augmented assistant to someone who might not be a trained mechanic.

CDO Magazine appreciates Duncan McCaskill for sharing his insights with our global community.