Why machines need our minds

The supremacy of the end user means humans are still needed to provide actionable insight.

Bookmark and Share

As information professionals we are painfully aware of the problems caused by the explosion of data – particularly of unstructured data.  The problems are not just of volume, but of veracity, velocity and variety. Can artificial intelligence (AI) provide solutions – and will AI eventually take over the role of the information professional? These questions were addressed by SLA Europe’s latest event Artificial Intelligence and the Information Professional – threat or opportunity?

Nick Bombourg (of FindOut) described how a machine learns by taking unstructured content, building assumptions, checking these and creating structural knowledge. This structured knowledge must have a purpose – to help support better decision-making, or to improve product– or indeed worker-effectiveness.

Marc Vollenweider of Evalueserve agrees.  The creation and design of information products has to start from a human angle.  Like information professionals, he and his team must get in the minds of their end users to truly understand what constitutes helpful –and actionable – insights. It is this part of the process that machines (as yet at least) can’t perform.   He has other learnings with positive messages for information professionals:

  • Customers truly value (and pay for) customisation
  • The 'so what' comes from human intervention
  • Deep understanding of users is what informs great products and services
  • Their customers really appreciate live chat with real humans
  • Relationships continue to be important
  • Humans can understand and provide nuanced advice and make 'judgement calls'

The ethics and politics of information and data will be increasingly important as will the information professional’s ability to manage and orchestrate information from multiple supply streams. 

The advantage AI has is the scale it can work at. But this accounts for work done at the lower end of the information value chain. For now, at least, machines still need us!