Two widely accepted beliefs about data and AI are that:
1) data is fundamental to AI development.
2) in order to ensure that AI is safe and responsible, we need to understand how that data has been collected and used across the AI lifecycle.
Sadly, these two beliefs that fit hand-in-glove have not led to widespread data transparency across the AI sector. Central to this disconnect – despite many documentation approaches being developed – is that there has been little research into what different potential users of this transparency information actually require.
This webinar, delivered by ODI researchers Ben Snaith and Sophia Worth in conversation with Professor Elena Simperl, details these research needs, findings from a project within the ODI’s Data Centric AI programme and shares the results of the first version of the AI Data Transparency Index which analysed 22 models.
About the event
You will be emailed the Zoom link.