The End of Expertise? Or A New Beginning?

Posted by:

Bill Fischer, Professor at International Institute for Management Development in Lausanne, Switzerland, recently penned a short op-ed nugget titled “The End of Expertise.” The piece is a “perspective” written to coincide with the 7th Global Drucker Forum being held in Vienna in November, and it was published on a no less respected platform than the Harvard Business Review website.

A major tenet of Fischer’s argument is that today’s decision makers need merely enlist “AI” to assemble from the Internet the answers to whatever questions are perplexing them. In this day and age of cognitive kinds of intelligence and oceans of content, he feels that decision makers will no longer have a need to seek out the kinds of for-hire professionals they used to bring on board to handle problems with deep domain expertise requirements.

Particularly because much of the content that documents the knowledge of the various professions is now published on the Internet (everything from complete academic curricula to professionally curated competency models, etc.), Fischer is concerned that many professionals (he mentions engineers, accountants, tax preparers, wine newsletter writers) may no longer be able to command the kinds of premium prices for their work to which they have become accustomed. Instead, their (presumably former) clients will simply step up to their iPhones and let Siri bring them the answers to their questions for the cost of a monthly service contract.

There is of course a great scare factor in the premise of this argument, particularly for the HBR audience, but no one should panic or lose any sleep. The basic premise—that “AI” will deliver “Credible” and “Reliable” high level knowledge—rests on an unsupported proposition that the Internet will deliver accurate, factual, contextualized, high quality “answers” to questions of the kind formerly posed to highly paid consultants and other professionals.

In fact, as even the most optimistic supporter of cognitive computing will tell you, today’s systems generally struggle to come up with answers to even rudimentary questions in a contextually appropriate and accurate manner. The industry poster child—IBM’s Jeopardy-playing Watson machine—could only operate in the highly specialized environment of that one game—and it took some five years for teams of IBM researchers to train it up to do even that.

The broader field that the Jeopardy Watson was operating in—“open domain question answering”—is still very much an area of active research in the computer science academy, and commercial examples of solutions are only beginning to emerge in highly constrained domains. Even the most advanced systems today stay away from claiming that they can deliver a correct answer (much less THE correct answer). Instead, they offer the human decision maker a short menu of candidate answers and leave it up to the human brain to choose one, or come up with a hybrid, or take another approach entirely.

So the vision of broad-based clairvoyant knowledge machines offering something like a general purpose iPhone app which will tell decision makers what to do in complex business situations is just that—a vision.

If you’re still worried about losing that high-paying consulting gig to Watson or Siri, I recommend that you try asking the Internet a question about something you really care about (like maybe about the best approach to some kind of heart surgery that you face or how to analyze the counterparty risk on your synthetic CDO) and see how that goes. In the meantime, both AI and the content it works on have a long way to go to stand up an advisory application that anybody could or should place much in the way of trust in.

We’re not approaching the end of expertise. In fact, we’re just beginning a phase where knowing just how to get the new cognitive tooling to take care of the heavy lifting that lays the foundation for successful decisions will become a new valued competence in and of itself.

Read Fischer’s End of Expertise at HBR


About the Author:

Hadley Reynolds is Co-founder and Managing Director at the Cognitive Computing Consortium. He is a leading analyst of the search, content management, and knowledge management industries, researching, speaking, and writing on emerging trends in these technologies and their impact on business practice. He currently leads the publications program at the Consortium.
  Related Posts

Add a Comment