Gartner Symposium – event review: Don’t just throw AI at it

09 November 2018 | AI
nicolas-comte-560698-unsplash

Gartner Symposium – event review: Don’t just throw AI at it

Having spent the last few days at Gartner’s European Symposium in Barcelona immersed in a culture frantically forecasting the technology and industry disruptions of the mid to long term future – it allowed me to take stock on the influencing technology trends of my own industry (RPA/IA) and the wider impact to white collar work. A thank you to Gartner for providing me with the opportunity to step away from the day to day and reflect on these wider trends.

Over three solid days it was hard not to find yourself in a seminar, briefing or corridor discussion that didn’t reference automation in some shape of form. Looking around the exhibition floor it was equally apparent that every product tagline is now infused with ‘AI’,  compounding confusion in the market around what Artificial Intelligence actually is, and more importantly, how it can be applied to the business in which an individual operates.

The sessions I attended stressed the need to empower organisations to embrace AI while being mindful of the social and ethical implications its use inevitably raises. To my own delight, the message of workforce ‘augmentation’ was a common thread in the advice shared on automation strategies. From my own experience, I see this position as a far more realistic long term viewpoint of the future architecture of the white collar workforce — rather than stories about the much-touted widespread displacement of FTEs from offices across world.

Aside from some thought-provoking use cases around pharmaceutical and healthcare AI initiatives, my overwhelming impression was that real live projects that showed a positive impact of AI technologies on business applications seemed few and far apart. An interesting statistic that emerged from a Gartner survey of EMEA executives, was that Chatbots were positioned as the top use case for AI – you can see the full breakdown of AI use cases below.  That seemed surprising to me given that Chatbots operating in isolation have a very limited business impact and are all too often being included as a checkbox activity, rather than organisation-wide benefit.

  • Chatbot 30%
  • Process Optimization 29%
  • Fraud Detection 21%
  • Market Segmentation 16%
  • Computer Assisted Diagnostics 14%
  • Call Centre Assistance 12%
  • Sentiment Analysis 11%

Even the categories noted in the survey reveal a definite confusion in use case segmentation.  For example, 'Sentiment Analysis' is more often than not a sub-component of Chatbots and 'Call Centre Assistance' solutions that are driven by part-automated conversational interfaces.  What does this tell us? It says to me that there is a lack of maturity in understanding how any of the AI subgroups can directly benefit the business world. Before launching into an AI programme, business leaders need step back and clarify their strategic goals. Realigning plans in order to  leverage the right technology, in the right context, in order to drive through to a successful conclusion.

With the term ‘fail-fast’ being used almost as much use as the term AI itself, organisations across industries understand that traditional approaches to product consumption, or product development are beginning to change. Given the increase in the maturity of consumer technologies and applications, user experience is becoming the core driving force in product. Maersk Shipping provided an interesting use case; a proprietary software application which Maersk had long used to track its own shipments and vessels, has now been developed into a commercial solution. Adopted globally by shipping and logistics firms to give visibility of cargo locations and allowing users to drill down into management information provided via an IoT service.  I was struck by the fact that in Maersk’s case, hardware and connectivity improvements, as well as the maturity of their application development, was as much to thank as any intelligent or cognitive services.  My point is that business value does not depend solely on AI – so conversely we must work harder to uncover how to translate AI capability into concrete operational benefits.

Whilst there is no doubt that a change in the accessibility of intelligent solutions is here, I’m fearful that organisations embark on an AI journey for the sake of following the pack will find it hard to show ROI.  As a technologist, it’s sometimes hard for me to admit that the technology is not the most important element. So, while the excitement about AI has set many CTOs creative juices flowing, I now feel an equal responsibility to my own firm and our clients, to shape and reference clear business use cases that provide a pragmatic approach to applying AI to business.

by Danny Major, CTO, Thoughtonomy