"To fully unlock the benefits of artificial intelligence, you’ll need to upgrade your people’s skills — and build an empowered, AI-savvy workforce."
Jeanne Ross MIT Sloane Mangement Review July 2017
Take fraud a frequently promoted use case for AI. In the rush to automation it is assumed that high volume lower value insurance claims fraud can be identified with AI.
Bitter experience shows that fraudsters are ahead of the algorithm authors and vary MO to beat the algorithms. Typically 30% to 35% will be fraudulent and will be caught with a combination of digitising the claims process and having claimants provide text and photo/image proof of damage.
And as the article below says-
"As other researchers and managers have also observed, we are finding that most machine learning applications augment, rather than replace, human efforts. In doing so, they demand changes in what people are doing. And in the case of AI — even more than was true with ERP systems — those changes eliminate many nonspecialized tasks and create skilled tasks that require good judgement and domain expertise."
Fix that flaw- creating skilled tasks and people, AI will be as disappointing as many ERP deployments.
For example, fraud detection applications may reduce the time that people spend looking for anomalies, but increase requirements for deciding what to do about those anomalies. An AI application might allow financial analysts to spend less time extracting data on financial performance, but it adds value only if someone spends more time considering the implications of that performance.
https://sloanreview.mit.edu/article/the-fundamental-flaw-in-ai-implementation/
unknownx500
