Ethical AID (Ethics in AI Development)
Lead Research Organisation:
University of the West of England
Department Name: Bristol Robotics Laboratory
Abstract
The great biologist E.O. Wilson once said, "The real problem of humanity is the following: we have Palaeolithic emotions, medieval institutions, and god-like technology." With AI and Robotics, our technology is becoming more truly God-like as we bring into being machines that can do some of the thinking for us, and share some of the properties that we have previously believed to be the exclusive domain of living creatures.
As this technology explodes into our world, it becomes more and more vital that we design systems that don't produce unintended consequences. This is not a questions of 'good' and 'bad'; conceptions of these vary between times, nations and cultures, and what is good in the context of a military system might be very inappropriate in a care home. Such moral framings vary, but the Ethical process of examining them and coming to judgements can remain the same, and could be reapplied.
It's easy, but unrewarding, to be wise in hindsight; what we need is ways of forecasting the ethical dimensions of a technological process, product or system as it is developed. This way companies might avoid wasting money on developing an unwanted project or rejected brand, and humanity might avoid finding out the hard way which technologies are best avoided.
The Ethical AID project will look at the way in which such an ethical forecasting system is implemented in one of the UK's powerhouse engineering companies. How has the investment been made to develop such a thing, how was it done, and what lessons can we learn from the process so that other companies might do something similar? Additionally, we will assess the effectiveness of the approach. To put it simply; how can we state that the world is a better place as a result of this process of ethical forecasting, and that we haven't just wasted our time in idle speculation?
We will then take our results to SMEs to test whether the insights we have gathered could be useful to them and provide the basis for further work beyond the project's end. In the long term, we need to build ethics into our process of technological development if our God-like powers are to make us a Heaven, not a Hell.
As this technology explodes into our world, it becomes more and more vital that we design systems that don't produce unintended consequences. This is not a questions of 'good' and 'bad'; conceptions of these vary between times, nations and cultures, and what is good in the context of a military system might be very inappropriate in a care home. Such moral framings vary, but the Ethical process of examining them and coming to judgements can remain the same, and could be reapplied.
It's easy, but unrewarding, to be wise in hindsight; what we need is ways of forecasting the ethical dimensions of a technological process, product or system as it is developed. This way companies might avoid wasting money on developing an unwanted project or rejected brand, and humanity might avoid finding out the hard way which technologies are best avoided.
The Ethical AID project will look at the way in which such an ethical forecasting system is implemented in one of the UK's powerhouse engineering companies. How has the investment been made to develop such a thing, how was it done, and what lessons can we learn from the process so that other companies might do something similar? Additionally, we will assess the effectiveness of the approach. To put it simply; how can we state that the world is a better place as a result of this process of ethical forecasting, and that we haven't just wasted our time in idle speculation?
We will then take our results to SMEs to test whether the insights we have gathered could be useful to them and provide the basis for further work beyond the project's end. In the long term, we need to build ethics into our process of technological development if our God-like powers are to make us a Heaven, not a Hell.
People |
ORCID iD |
| Matthew Studley (Principal Investigator) |
| Description | I have developed a new research network which brings together key staff from across Arts and Humanities, Engineering and Computing, and Robotics. Members are mainly from UK academia, with some involvement from European resarchers and partners from Industry. At our first meeting we had a day of wide-ranging discussion to frame questions about the ways in which Arts and Humanities could shape our AI future. We are following on from this workshop with publications and reporting, and we have gained insights into ways we could involve more industrial partners at C-Suite level in an engagement beyond the normal process focus of "ethics in support of technology". |
| Exploitation Route | I would welcome contact from researchers engaged in ethics and AI, especially from an Arts and Humanities perspective, who could gain value from joining a wider network. |
| Sectors | Aerospace, Defence and Marine,Digital/Communication/Information Technologies (including Software),Government, Democracy and Justice,Security and Diplomacy |
| Description | Collaboration with Prof Scott Delahunta |
| Organisation | Coventry University |
| Country | United Kingdom |
| Sector | Academic/University |
| PI Contribution | Network development and thought leadership. Funding for T&S and venue for workshop held at RA Eng on 27th Jan, 2023. |
| Collaborator Contribution | Network development and thought leadership. |
| Impact | Scott Delahunta leads a centre for Dance Research at Coventry. We started to collaborate as a result of my growing belief that ethical AI development requires the living practice of ethics, over and above the duty to external standards which is starting to drive the field. We co-organised a workshop and developed a network which we continue to nurture, and will report upon. |
| Start Year | 2022 |
| Description | Multidisciplinary workshop |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Professional Practitioners |
| Results and Impact | I organised and co-hosted a workshop entitled, "How can the Arts and Humanities shape our AI future?", working with contacts I have developed through talks delivered as part of this award. We had over 20 attendees from UK and European Universities, The Ada Lovelace and Alan Turing Institutes, AHRC / UKRI, and from varied disciplines and backgrounds. We are engaging this network with post-work and have been invited to report on it in the Journal 'AI & Society'. |
| Year(s) Of Engagement Activity | 2023 |