- Only slightly more than half of the respondents to GlobalData’s recent survey on emerging technologies felt that they fully understood artificial intelligence.
- Before decision-makers act on the insights revealed by artificial intelligence, they need to have confidence in its findings, which requires an understanding of how they are obtained.
There is a widely accepted belief that artificial intelligence (AI) holds the potential to significantly alter the way organizations operate, vastly improving business outcomes. Businesses conceptually grasp that the technology’s benefits range from increasing efficiency and productivity to enhancing the customer experience. Yet, AI is still often viewed as a mysterious black box that yields little insight into how its findings are obtained. Without understanding what is happening under the covers, line-of-business leaders can be reluctant to act on the findings.
As shown below, only slightly more than half of the respondents to GlobalData’s recent survey on emerging technologies felt that they fully understood artificial intelligence. Close to 40% noted that they partially understood the technology; the remaining 6% indicated that they didn’t know anything about artificial intelligence.
Several AI platform providers are addressing this challenge by promoting ‘explainable AI’ tools. These tools reveal the factors that contribute to an AI model’s findings and provide a numerical score to convey the degree to which a data point influences results.
Lack of model understanding is just one of numerous challenges impeding business adoption of artificial intelligence, which can range from a lack of resources to concerns related to unintended bias in models. However, as with other challenges, it is one which is being addressed with new tools and capabilities that will undoubtedly become easier to use and more prevalent over time.