Our Engineering Responsible AI project is grounded in the realities of AI technology and presents the balanced views of experts to unpick the polarisation and exaggeration in the debate. Through work which is specific, problem-focused and policy-driven, we seek to provide advice and recommendations to help the UK enjoy economic and social benefits from the development of safe, secure and sustainable AI.
The People's AI Stewardship Summits
The People's AI Stewardship Summits are a series of public dialogue events bringing together members of the public with experts, innovators, entrepreneurs and other stakeholders to explore hopes, fears and expectations of AI to enable positive change and improve people's lives. The events are held in partnership with, and co-hosted by, the Academy's Regional Enterprise Hubs to create a unique space for stakeholders to listen to local members of the communities in which they are based. We held summits across five UK cities Belfast, Glasgow, Swansea, Liverpool and London.
A series of vox pop videos were produced at each summits asking members of the public to respond to questions such as:
- How could you trust AI?
- How do you think AI could benefit Belfast?
- Do the public understand AI?
Find out more about the AI summits
New event series Practical AI: Inform
Practical AI: Inform is a new talk series discussing practical AI adoption. Hosted by Academy President Sir John Lazar, we’ll be hearing from adopters, users and AI developers about their experiences and the practical lessons they’ve learned.
Al taxonomy
In order to provide a common language for discussing the key facets of AI, Academy Chair in Emerging Technologies, Professor Themis Prodromakis worked with us to develop an AI taxonomy. This can enable researchers, developers, and policymakers to understand where they are intervening within the wider AI system and encourages consideration of the implications of design decisions at other layers.
Watch the video below to explore the taxonomy in greater detail.
Video transcript
[Music]
[AI Taxonomy title on screen]
Artificial intelligence technology is growing at an astonishing pace giving rise to transformative Innovations. However AI is not a homogeneous mass but a multifaceted kaleidoscope of subfields to facilitate communication between innovators, investors policy makers and the public. The Centre for Electronics Frontiers and the Royal Academy of Engineering have created a five layer taxonomy for navigating AI. At the base is the hardware layer the components that form the physical building blocks of an AI system. Innovations in semiconductor materials and devices here can maximize the speed and efficiency of an AI system. Next is the functional layer this includes all fundamental mathematical functions that underpin the systems computation from simple pocket calculator style processing to collections of artificial neurons that can render biologically inspired processing capabilities. The third computational layer is all about the labelling of data creating a vocabulary of symbols that a machine can use to identify and understand meaningful units within images speech or other data. In the fourth semantic layer machines can predict how how complex systems might evolve and expand their previous learning to be able to deal with completely new situations. The final agency layer is the part of AI that enables machines to understand goals and evaluate their actions. This is where capability meets intention, giving rise to key ethical dilemmas of artificial intelligence. It is also where autonomy reaches its highest capabilities. Naturally all the layers are tightly interconnected and the boundaries between them are blurred but the use of this taxonomy can provide a better understanding on how innovations in any layer can bring significant improvements to others. Creating a natural innovation chain ecosystem, improving collaboration across layers and bringing together researchers from different disciplines leading to even greater AI innovation.
[cef.eng.ed.ac.uk appears on screen]
Recent AI policy work
Foundations for environmentally sustainable AI
A new report from the NEPC highlighting foundational steps and key actions to minimise risks to the environment, people…
Glasgow
The second People’s AI Stewardship Summit took place in Glasgow. This summit focused on exploring the themes of AI in h…
Data and AI
The Academy has undertaken a wide range of projects on the role of data and artificial intelligence (AI) in shaping the…
Safety and ethics of autonomous systems
Exploring how autonomous systems should be designed, developed and deployed to ensure widespread benefits.
Latest news and blogs
More news
UK Government urged to promote, prioritise and invest in sustainable AI to become global leader in AI frugality and efficiency
New report calls for extension of mandatory reporting on AI’s energy and water use, carbon emissions and e-waste recycl…
Quantum technology could open the way to economic advantage for the UK, says Academy
The UK is well placed to apply its unique expertise and commercial knowhow in key areas of quantum technology that coul…
Engineering Responsible AI: Q&A with Charlette N’Guessan
The winner of the 2020 Africa Prize shares her thoughts on how generative AI can be more inclusive and questions the ps…
Academy comments on new AI white paper
Professor Nick Jennings CB FREng FRS, Chair of the Engineering Policy Centre Committee, comments on the new AI white pa…
Explore more of the Academy's work on AI
Data and AI
The Academy has undertaken a wide range of projects on the role of data and artificial intelligence (AI) in shaping the…
Safety and ethics of autonomous systems
Exploring how autonomous systems should be designed, developed and deployed to ensure widespread benefits.
Critical capabilities
Effective responses to a national emergency requires the ability to call on and rapidly deploy critical capabilities: t…