A clear and persuasive analysis in the Tuesday (March 5) edition of the Financial Times (“40% of Europe’s artificial intelligence start-ups have no AI”) illustrates the high levels of misplaced expectations around the short-term benefits deriving from artificial intelligence, machine learning or data science.
The main cause of the current misplaced high expectations is due to oversimplifications in public media and marketing materials, as well as the general lack of interest in the actual mathematical methods relative to the applications of machine learning.
Irrespective of your choice of the favorite label (AI, machine learning or data science), this domain does represent significant business opportunities across industries as well as across work functions. The devil is in the detail as each business needs to assess how to best position itself in the context of its business aspirations.
Machine learning, as I prefer to call it, is a general purpose technology [1], and represents a structural game changer. It facilitates higher levels of transparency and evidence-based decision-making, as well as automation.
In short, the combination of data and statistical models stands to benefit all types of businesses. However, the adoption of these technologies will take years (if not decades), as was the case with electricity in the early 20th century [2], and superior seeds of hybrid corn [3].
Article links: https://www.ft.com/content/21b19010-3e9f-11e9-b896-fe36ec32aece.
[1] https://en.wikipedia.org/wiki/Electrification
[2] https://www.technologyreview.com/s/612897/this-is-why-ai-has-yet-to-reshape-most-businesses/