As part of this assignment, we were asked to review a news article from a major online publication whose topic is related to statistical tools and the Central Limit Theorem in particular. Such an article was the March 2021 piece chosen, namely Three Key Conditions For Robust ML Solutions by Prashanth Southekal. The article was published under the Forbes label, which means that the publisher’s authority imposes some criteria, allowing less doubt about the reliability of the published data.
In brief, this material describes the tendency of modern enterprises to use AI and ML as tools for optimizing operational efficiency. The author gives a pretty accessible explanation of these concepts (given their relative academic novelty), followed by excerpts from a statistical report by McKinsey & Co. For example, Southekal writes that the introduction of AI in corporate practices increases profits by an average of 79%. At the same time, for companies with 20% of profits from AI, the speed of launching new projects increases because of remote working opportunities. The author then argues that 80% of all data collected in the enterprise is unstructured, and CLT should be implemented to optimize machine calculations. More specifically, Southekal cites information without reference when he says that the number of records to process must exceed 100 to comply with the CLT concept. As a result, the author cites three critical criteria for using machine data based on the report data and provides a set of three recommendations for companies deciding to integrate AI and ML into their business practices.
Referring to the source makes it clear that in his article Southekal referred only to the data that were most relevant in the context of the topic: the report provides additional statistical information. However, there is some inaccuracy in the management of the numbers. From the article, it seems that companies that get 20% of their earnings from AI launch new projects faster, while the report indicates “…a small contingent of respondents <…> attribute 20 percent or more of their organizations’ earnings <…> to invest even more in AI” (McKinsey Analytics 2). In addition, the article mentions a 79% increase in profitability of companies but does not indicate that there is a trend of a 25% reduction in costs. Thus, we can determine that the news article does not accurately reflect the statistics, while the overall message can be modified.
References
McKinsey Analytics. (2020). Global survey: The state of AI in 2020 [PDFdocument].
Southekal, P. (2021). Three key conditions for robust ML solutions. Forbes.