AI has once again become “cool new tech”, an exciting new way to address old challenges. We have largely been taken by surprise by the tremendous advances in capability of the latest Large Language Models (LLMs) and Generative AI. These new data driven, algorithmic tools have redrawn the frontier of what we thought Artificial Intelligence (AI) could do.  

In the past, we have used AI to automate, navigate, detect anomalies, recommend next actions, patten match, predict and explore “what if” simulation scenarios. The newer AI can do that and generate, synthesise, translate and intelligently tackle moderate complexity tasks. They can do this whist being judged to “demonstrate” greater empathy and patience than human respondents in online environments constrained to text-based interactions.  

What has changed is the way AI works and the size of the datasets used to train the AI. Generative AI is trained to “focus” and is training on datasets of unimaginable sizes to mere mortals, literally trillions of examples. This unsupervised training occasionally leads to some surprises. When presented with a supposedly factual response from your AI query, some results may refer to “real world” sources that simply do not exist. This scaled up version of an age-old adage of “garbage-in-garbage-out” leads to the modern twist “garbage-in-sometimes-hallucination-out”.  Nonetheless, the responses from the latest generation AI tools, are pretty impressive, even if they need to be fact checked.  

The reality is that AI is a use of data, a powerful one, but still one that relies on understanding data quality, provenance, fitness for purpose and some seemingly unexciting elements. If the data is not fit for the purpose that it is being used for, then there is no hope of getting something useful from the AI.  Similarly, it is essential to understand how much you can rely on the output from your AI.  

This workshop explores some historical and current uses of sophisticated AI. It explores some of the most recent types of AI and the associated challenges associated with it. It takes you just below the surface of AI to explore what a LLM is, and if time, some other classes of AI algorithm. It will also delve into data sharing frameworks which can be used with AI and gives an overview of an AI Assurance framework to help make AI really useful.