Responsible Engineering, Design & Innovation (REDI)

Which way forward?

 

With the advent of technology that can learn and change itself, and the integration of vast data sources tracking every detail of human lives, engineering now entails decision-making with complex moral implications and global impact.  As part of daily practice, technologists face values-laden tensions concerning privacy, justice, transparency, wellbeing, human rights, and questions that strike at the very nature of what it is to be human.

We recently edited a Special Issue of IEEE Transaction on Technology and Society about “After Covid-19: Crises, Ethics, and Socio-Technical Change”

"Our research works to understand the paths toward a future in which technology benefits all of humankind and the planet. We collaborate with social scientists to develop practical methods and socio-technical solutions to equip engineers and designers with the tools necessary for practicing responsibly through every step of the development process. "

Partners

Centre for the Future of Intelligence

We are core members of the Leverhulme Centre for the Future of Intelligence, an AI Ethics research centre bringing together world leading AI researchers in technology and the social sciences and based at Imperial College London, University of Cambridge, University of Oxford and University of California, Berkeley.

Go to website

Digital Catapult

Rafael Calvo is a member of Digital Catapult's 'Machine intelligence Garage' Ethics Committee, which was created to convene experts in AI and data ethics to 1. oversee the development of principles and tools to facilitate responsible AI in practice, and 2. work closely with startups developing through Digital Catapult’s Machine Intelligence Garage programme.

Go to website

Projects

UNESCO - The Future of Work

Professor Rafael Calvo is working with an international team of researchers in industrial relations, psychology, history, education and sociology on advising UNESCO in relation to the future of education and work in a world of AI and uncertainty.

Report

AI, Bias & Disability

New technologies can advance the lives of people with disabilities, but evidence is emerging that some AI may be entrenching bias or even generating new forms of injustice. For the research project "Designing AI to stop disability bias" we have partnered with colleagues at NTU in Singapore to ensure responsible AI in relation to people with disabilities.

Project Website

Responsible AI in the workplace

We work on designing technologies in ways that support precisely these factors that fuel motivation, engagement and wellbeing and we have methods for conducting design and research together with the people impacted. Our experience in doing so has made it clear that without addressing this human psychological layer, even the cleverest technology can fail.

Find out more

Culturally Appropriate Language Models and Systems (CALMS)

The Imperial team is working together with rural communities across four distinct regions of Peru to co-design responsible technologies that can help improve dementia care. This includes creating a dementia screening tool and an intervention program to support informal carers. The aim is to develop effective, accessible, and culturally sensitive solutions that can address the unique challenges posed by dementia in low- and middle-income countries.

Find out more

Responsible Tech Design Library

Find out more about tools and methods for more ethical practice in technology design

Staff

Prof. Rafael Calvo

Rafael Calvo

Prof. Rafael Calvo

Dr Celine Mougenot

Celine Mougenot

Dr Celine Mougenot

Prof. Sebastian Deterding

Prof. Sebastian Deterding

Dr Fangzhou You

Dr Fangzhou You

Laura Moradbakhti

Laura Moradbakhti

Dr Juan Pablo Bermudez

Juan Pablo Bermudez

Dr Juan Pablo Bermudez

Marco Da Re

Marco Da Re

Citation

BibTex format

@inproceedings{Widjaya:2023,
author = {Widjaya, MA and Bermudez, J and Moradbakhti, L and Calvo, R},
title = {Drivers of trust in generative AI-powered voice assistants: the role of references},
url = {http://hdl.handle.net/10044/1/106752},
year = {2023}
}
Download

RIS format (EndNote, RefMan)

TY  - CPAPER
AB  - The boom in generative artificial intelligence (AI) and continuing growth of Voice Assistants (VAs) suggests their trajectories will converge. This conjecture aligns with the development of AI-driven conversational agents, aiming to utilise advance natural language processing (NLP) methods to enhance the capabilities of voice assistants. However, design guidelines for VAs prioritise maximum efficiency by advocating for the use of concise answers. This poses a conflict with the challenges around generative AI, such as inaccuracies and misinterpretation, as shorter responses may not adequately provide users with meaningful information. AI-VA systems can adapt drivers of trust formation, such as references and authorship, to improve credibility. A better understanding of user behaviour when using the system is needed to develop revised design recommendations for AI-powered VA systems. This paper reports an online survey of 256 participants residing in the U.K. and nine follow-up interviews, where user behaviour is investigated to identify drivers of trust in the context of obtaining digital information from a generative AI-based VA system. Adding references is promising as a tool for increasing trust in systems producing text, yet we found no evidence that the inclusion of references in a VA response contributed towards the perceived reliability or trust towards the system. We examine further variables driving user trust in AI-powered VA systems.
AU  - Widjaya,MA
AU  - Bermudez,J
AU  - Moradbakhti,L
AU  - Calvo,R
PY  - 2023///
TI  - Drivers of trust in generative AI-powered voice assistants: the role of references
UR  - http://hdl.handle.net/10044/1/106752
ER  -
Download