Three years ago, I was lucky to be asked to product lead a key initiative for my clients British Council, key leaders in the English teaching and assessment: help define the future of teaching learning and assessment. Back then, when we started, 'AI', as it is ubiquitously seen today wasn't quite around, so when the latest flavour of LLMs became available we needed to carefully decide whether this (and 'AI' in general) was technically and ethically the right thing, and how we would best use 'AI' capabilities, in this highly sensitive context.
In this talk I'll share how my team went about making these choices, considering technical and ethical trade offs. I will touch on general considerations such as quality and reliability, but also explainability, sustainability and bias.
Participants will not only see a cutting edge case study from an innovation team in a £1B global organisation, but also take away practical tips and frameworks to help them build ethical and sustainable 'AI' driven products and services.