Artificial Intelligence and the Information Professional: Threat or Opportunity?
Our thanks to Richard Nelsson, Information Manager at Guardian News & Media, who kindly accepted to offer a review of the event on artificial intelligence & the information professional, which took place last month in London.
The world is rapidly moving towards a fourth industrial revolution. The rise of robotics, artificial intelligence (AI) and other technologies will see machines that ‘think’ like humans not only eliminate repetitive tasks but could also help solve huge problems such as finding a cure for cancer.
This change though will bring severe disruption to the workplace. According to a recent World Economic Forum report, it will lead to a global loss of five million jobs by 2020 – many of them knowledge workers in professions such as law and medicine.
Of course for information professionals working with ever-increasing volumes of unstructured data, AI and big data analytics offer welcome solutions. But at what cost? These issues were addressed at a packed SLA discussion, Artificial Intelligence and the information professional: threat or opportunity?
Nicolas Bombourg, managing director and co-founder of Findout, began by stating that the purpose of AI technology is to provide tools for making better and faster business decisions, as well as product and worker effectiveness.
At the moment, 80% of data is unstructured. This could be in a non-machine readable format, buried, or part of a system that is constantly updated and changed.
AI gives structure to this data by building assumptions or predictive models, making different technologies work together. The system then has the analytical abilities to make decisions on its own. Ultimately though, humans still have a key role to play in this workflow.
Continuing with this theme, Marc Vollenweider, CEO and co-founder of Evalueserve stated that humans have to be at the heart of any AI system. While mind-only analytics, where humans do all of the work is too slow and too costly, machine-only analytics rarely delivers the qualitative insights needed. The human mind is required to make ‘judgment calls’ – the right insight delivered in the right format at the right time. His team attempts to understand what the user requires, which then informs the products and services developed. Live chatting with humans is also important – as demonstrated by Marc connecting with a colleague in South America. This was all very encouraging news – at least for now.
Questions, chaired by John Coll, SLA Europe President, came thick and fast, no doubt encouraged by the offer of a free copy of Marc’s new book – mind+machine – for the best three. Areas covered ranged from copyright and an agency’s decision to block sharing ‘juicy’ data, to the ethical issues surrounding AI. Guarding against mistakes and bias is particularly important. Here, the recent example of Volkswagen (a sophisticated software algorithm was developed to ensure exhaust emissions would always meet government regulations) was given to show that technology misuse is essentially fraud.
John then invited everyone to continue the conversation over canapés and drinks in the bar. Chatting over dinner – ‘canapés’ turned out to be a full buffet – the general consensus appeared to be that it had been a fascinating evening, but a few more practical examples of data tools in action would have been welcome.
Hats off to the organisers though for putting together a useful event that highlighted the changes and challenges presented by AI. It is an area SLA will be sure to return to over the next few years. Thanks too to the sponsors Findout and Evalueserve and to Morgan Stanley for hosting.