"In many practical settings, AI tends to be most successful as a tool to augment rather than replace decision-making."
Julia Handl explains why AMBS is developing increasing links with companies in the field of decision sciences.
The recent announcement that AMBS now has five Alan Turing Institute (ATI) fellows speaks volumes for our strength in data sciences.
The UK’s National Institute for Artificial Intelligence recently announced more than 400 fellowships for 2021/22 across its 12 university partners, with The University of Manchester now home to 33 fellows. At AMBS Dr Richard Allmendinger, Professor Yu-wang Chen, and Dr Manuel Lopez-Ibáñez were all appointed new fellows for 2021/22, while myself and Dr Nadia Papamichail had our fellowships renewed.
These fellowships further cement the increasing links that we are developing with companies in the field of decision and sciences, areas which are becoming increasingly important to them as they seek to analyse and make use of the mass of data that flows through their businesses hour by hour, day by day.
Part of our role when talking to businesses is to explain the practical constraints of AI (Artificial Intelligence) and to discuss some of the common misconceptions surrounding the use of AI and machine learning.
For instance, it is not uncommon for individuals to equate AI and machine learning, when it might be more appropriate to think of machine learning as just one subset of AI. Even in data-rich environments, certain problems may be more suitably tackled using traditional operations research such as simulation and optimisation techniques. In many practical settings, AI tends to be most successful as a tool to augment rather than replace decision-making, so consideration needs to be given to aspects of human-computer interaction and the design of effective decision support tools.
One area where we are seeing increasing activity with business is around Knowledge Transfer Partnerships (KTPs), which are a really effective model for collaborations in the field of AI. A successful AI project depends on both strong technical skills and excellent domain insight, and KTPs achieve this by sending technically skilled experts to be embedded directly within the relevant teams of a company.
Also, we are finding that our fellowships with the ATI have great value for companies, as they open up an additional range of opportunities for engagement and joint funding applications and increase the visibility of the work.
There is also a really strong internal talent development angle to many of our collaborations with companies. KTPs are designed specifically to help both with the translation of state of the art knowledge into practical industry problems, and the transfer of the associated knowledge to teams within the company.
Where MSc and PhD students are involved in collaborations this brings tangible benefits for both sides. Companies and organisations often go on to recruit our students directly, benefitting from this first-in-line access to our graduates. Our students gain practical experience in the translation of their technical skills, valuable work experience and, potentially, a direct route into their first job.
Another area where our work is gaining greater traction is in how AI is now embedded into a greater range of teaching programmes. Our Business Analytics programme is designed to equip students with a broad and powerful toolkit of techniques ranging from operations research over data science to decision-making. AMBS also looks after the Business and Management pathway of the Data Science MSc, The University of Manchester’s cross-faculty flagship programme in Data Science.
At AMBS we recognise that data literacy and the interpretation of data are becoming essential skills, and there isn’t one area of our business school that isn’t touched by this agenda. At the end of the day, being able to make sense of data is really important, not only to arrive at good decisions yourself but also to assist with interpreting information provided by others.
This article was originally published in issue 9 of the AMBS Magazine.