The underlying article provides some analysis to potential benefits and pitfalls of using AI software to identify future musical hits and stars.  More widely, however, it allows for reflection into the role of such technology in shaping the manner in which goods and services are sourced.

Whilst some level of divergence from initial parameters can be programmed-in, fundamentally "AI can’t learn something that it hasn’t been taught", meaning that individuality and innovation (arguably of most importance in creative fields, given the subjectivity of taste) may be overlooked and stifled if we are simply relying on discernment (and attribution of worth) based on prior experience.   

Our technological memory banks (borne from the choices that we've previously made through technological means) can have an impact on what we are fed (including quite literally, as e.g. we may be directed to restaurants/takeaways based on past behaviour). This may explain why some feel that AI is an imperfect substitute for human selectivity and rather should be carefully employed and managed to add efficiency in the choices that we ultimately make.  

Nevertheless, given the inescapable burgeoning of this technology, and since the same can impact upon the whole spectrum of goods and service industries in directing choice, as legal and other service providers it has never been more important to ensure that our net (in the form of the technological footprint of our expertise) is cast far and wide.