Even though artificial intelligence (AI) and deep-learning techniques will be the most common approaches for new applications of data science, these new technologies come with their own set of challenges, the leading one being peoples’ suspicions of analysis they don’t understand.
Earlier, analytic techniques were always comprehensible, at least in conceptual terms. Even if users did not understand a specific model, they understood the logic behind it.
Deep learning, even though it delivers more-accurate results, lacks transparency. If users are unable to comprehend how the solution was achieved, they tend to distrust the technology.
Another concerning restriction on the corporate adoption of deep-learning techniques is the threat of lawsuits over difficult-to-explain algorithms that could make a mistake. D&A leaders should involve the legal department before deploying a new technique.