While AI (Artificial Intelligence) seems to be everywhere, amidst the hoopla over stars such as Alexa, Alpha Go or Siri, the technology still has a lot to improvise. While innovators say that AI will soon be a part of almost everything, it is better to keep a cautious eye out. As per a recent report by the McKinsey Global Institute, even within sectors, there remains a massive disparity between the leaders and followers. The sectors on top of the AI adoption rates are financial services, high tech, communications, logistics, healthcare and tourism. This is followed by a middle pack comprising media, professional services, retail, energy, education, automotive and consumer packaged goods. Building materials and construction industry is falling behind. Five major limitations have been identified by this study. The first of them is, data labelling, which is the process of each data set being categorized by humans. This leads to a reinforced learning using trial and error. Another limitation is the requirement of vast data warehousing to get any operation underway. A third is that while AI provides quantitative analysis, it does not explain the bigger picture. AI also does not have the ability to improvise in learning, as rather it is too generalized. Ideally, AI is supposed to be free of biases, but ultimately all algorithms are created by humans, so inherent biases creep in. A calibrated approach with lateral thinking is needed to solve this, with a sophisticated strategy for business analytics.

Source:https://www.mckinsey.com/business-functions/mckinsey-analytics/our-insights/what-ai-can-and-cant-do-yet-for-your-business?cid=other-soc-twi-mip-mck-oth-1801&kui=1UXNVzIzR8YwtGkEDN4iUQ

Uploaded Date:13 February 2018