Resolving the AI paradox – Advice for Data Scientists who want to advance the world

Companies are discovering that it is fiendishly hard to move from successful Artificial Intelligence (AI) prototypes towards AI at scale. We call this the “AI paradox”, and we want to share our experiences on why data science user cases often are hard to scale, and on what it takes to successfully go from prototype to an industrialized implementation.

From examples from our own work with some of the world’s largest companies and most innovative AI projects, we will explain why AI does not scale the same way as traditional IT systems, why you should always be wary about proof-of-concepts, and why the most important integration test you do is when integrating business people on your team.