Harker at the Liverpool City Region AI Summit

Tom Houghton-Alliss (Graduate Software Developer) writes…

Myself and the Harker team attended the first ever Liverpool City Region (LCR) AI Summit! It was an interesting day filled with panels and discussions, all about how AI is shaping the future. We learned about the vast benefits of AI, as well as the current concerns with using and training it - as well as hearing from professionals across the industry!

The event opened with a keynote from Nicola Hodson, Chief Executive of IBM, who gave an overview about AI in business and how it will go on to affect our daily lives. This was followed by a number of panels featuring individuals representing a wide range of industries. We had doctors, lawyers, small tech start-ups, video game developers, and even individuals in the public sector give their take on AI and its role in ethics, legality, trust, innovation, and digital inclusion.

The benefits of AI are already being felt in our most critical services. We have doctors able to have more personal appointments with patients, while an AI takes details notes on conversations and interactions. In the fire service we have AI identifying at-risk households, people who need more assistance when it comes to fire prevention. In schools we have AI quickly identifying gaps in a child’s understanding of topics, and providing support and information to fill those gaps.

Issues arise as we consider AI and people - for example, how can we harness the power of AI and use it as a tool to support employees, rather than replace them? Or, how can we ensure that people using AI know to check for mistakes and misinformation, rather than blindly trusting its output? How can we ensure that the data being fed to these Large Language Models (LLMs) is normalised and high quality - to prevent polluting its outputs? These are all problems that must be tackled before AI can be widely adopted.

I think the biggest issue facing AI at the moment, the one blocking it from enhancing our critical decision making, is that it is widely “black box” - this means an AI cannot explain to a user exactly how it came to its decision - that is, AI is not “explainable”. For example, if we pass patient data to an advanced AI and task it with selecting the better of two high risk surgical procedures, we need to know exactly what data was considered, how it came to the final decision, and specifically what training data was referenced in making that decision. We need to know if the AI decided to ignore key pieces of data, or if its internal logic is faulty, or if it considers some of the training data as indecipherable. Without knowing all of this and more, we can’t put very much stock in the decisions this AI will go onto make.

After these panels concluded, we had closing remarks from Katherine Fairclough, Chief Executive of the Liverpool City Region Combined Authority, who discussed AI and its implications in the public sector - for example, could we use AI to enhance public transport? If so, what would the training data look like for doing so? And what sort of impact would we hope to get from harnessing AI in this context? These are questions that are currently unanswered, and provide a challenging problem to solve for the future.

And finally, we had a closing speech from Steve Rotheram, Metro Mayor of the Liverpool City Region, who talked about how proud he was that Liverpool was a home to these discussions, and how he had high hopes for the continued innovation and improvements to AI and its societal impact.

At Harker, we’ve taken all of these talking points on board in our development process, and we’re excited to be developing an AI-enabled platform within the merging hotspot of “AI for good” - where we aim to harness the power of AI to boost social benefit and improve the lives of the people around us.