The use of generative AI is becoming increasingly widespread across a growing number of fields. Taking knowledge work as an example, this multiplier case study used the Friendly Tech Check to identify the potential benefits and risks that members of a development team at a major research institution perceive in relation to their work. Furthermore, discussions with staff led to the development of ideas for a human-centred approach to its use.

During a staff dialogue with members of a development team, fundamental application scenarios and design requirements for generative AI in knowledge work were discussed. The Friendly Tech Check was then used to identify the specific opportunities and risks associated with the use of ChatGPT in knowledge work, as perceived by the development team. Following this, requirements for a human-centred use of ChatGPT were derived, and the results were fed back to the employees in a workshop.

The Friendly Tech Check revealed mixed effects of generative AI in knowledge work. Alongside its potential benefits (e.g. relieving staff of routine tasks), risks also became apparent (e.g. so-called ‘hallucinations’). From the employees’ perspective, there are three key areas where policy measures are required: (1) raising organisational awareness, (2) training and skills development programmes, and (3) knowledge transfer and exchange between ‘early adopters’ and potential users.

Institute for Work and Qualifications, University of Duisburg-Essen