The trajectory of its future development cannot be known but it is clear that artificial intelligence will become increasingly embedded in the home and beyond. The complexities of AI mean it will have varied impact across different sectors of the economy and on their respective pathways of growth or decline. Jobs that are repetitive in nature and/or work towards binary outcomes are particularly at risk. Managing transformational AI will therefore require a degree of wisdom, understanding and individual responsibility on the part of programmers and end-users. At present there is a combined effort in establishing a permanent feedback loop that identifies and rectifies mistakes as the program is used (ChatGPT, for instance, utilises such a system).
The prospect of superintelligent AI raises ontological questions about what it means to be human. Judaeo-Christian teaching emphasises the implications of human beings as bearers of the imago Dei, and the role and responsibility this carries vis-à-vis the rest of creation. It therefore offers a unique perspective in developing AI in which the human being remains central.
Chapter 1 touched on key milestones in AI history, highlighting the gradual advancement from basic computational AI to such new technologies as machine learning, pattern recognition and deep learning. Chapter 2 presented two case studies: an analysis of AI-led biometric identification, which brought significant efficiency and security benefits to points of transit; and the development of autonomous vehicles which, though innovative, are as yet unsafe for widespread public use. Chapter 3 considered a Judaeo-Christian response to the adoption and use of AI. It emphasised the reality of humanity as bearing the image of the divine and the implications this has for sentience as a uniquely human attribute.
This concluding chapter offers three guiding principles for AI’s adoption and use.
First and foremost, AI has to be utilised and designed in a manner that holistically benefits humanity. Stuart Russell and Peter Norvig have written: ‘Given that AI is a powerful technology, we have a moral obligation to use it well, to promote the positive aspects and avoid or mitigate the negative ones.’[1] This is where early adopters and developers of new integrated AI technologies need to dig deeper. The goal should be societal value in forms that enrich relationships and communities, and support economic growth leading towards alleviation of poverty. From the Judaeo-Christian perspective, recognising what distinguishes humans from the rest of creation can act as a guide and catalyst for discerning a right path at a time when the detailed ramifications of AI remain unclear.
The second principle is to ensure that new technologies that can have an impact on the general population are rooted in a set of underlying ethical values. These would generally include such widely held beliefs as human dignity, the respect and protection of life, freedom of expression, fairness in the treatment of others, non-maleficence and justice. The integration of moral values into the AI code should come through, and in some cases from, the computer developers themselves. Developers have a moral duty to ensure to the best of their ability that harmful, unintended consequences are curtailed, and safety systems in place when things go wrong. Some tech companies have already taken concrete steps to implement ethics into their own AI products, and have helped others follow suit. IBM, for instance, launched AI Fairness 360, which is an open-source toolkit available to all developers.[2]
The third principle is that the regulatory framework surrounding AI should focus on the tenets of AI design rather than the intricacies. Though built on a set of principles, the process of innovation must be protected, and corrected at the point at which it fails – innovation should be welcomed with cautious optimism, not cynical suspicion. Adam Thierer, a technology policy expert at R Street Institute, argues against the excessive use of the ‘precautionary principle’ and points out that: ‘If public policy is guided at every turn by fear of hypothetical worst-case scenarios and the precautionary mindset, then innovation becomes less likely.’[3] Policymakers must strive to establish an environment that protects and harnesses the innovative spirit of technology entrepreneurs.
UK legislative proposals and initiatives affecting AI are currently embryonic. A report by the House of Lords Select Committee on Artificial Intelligence acknowledged in 2018 that while ‘AI-specific regulation is not appropriate at this stage’, the government should create an AI Code that serves as a shared ethical framework.[4] Such a framework would be based on five key principles:
- Artificial intelligence should be developed for the common good and benefit of humanity.
- Artificial intelligence should operate on principles of intelligibility and fairness.
- Artificial intelligence should not be used to diminish the data rights or privacy of individuals, families or communities.
- All citizens have the right to be educated to enable them to flourish mentally, emotionally and economically alongside artificial intelligence.
- The autonomous power to hurt, destroy or deceive human beings should never be vested in artificial intelligence.[5]
In 2020, a subsequent paper by the House of Lords Liaison Committee argued that the Centre for Data Ethics and Innovation within the Department for Science, Innovation and Technology:
should establish and publish national standards for the ethical development and deployment of AI. National standards will provide an ingrained approach to ethical AI, and ensure consistency and clarity on the practical standards expected for the companies developing AI, the businesses applying AI, and the consumers using AI. These standards should consist of two frameworks, one for the ethical development of AI, including issues of prejudice and bias, and the other for the ethical use of AI by policymakers and businesses. These two frameworks should reflect the different risks and considerations at each stage of AI use.[6]
Vague and broad statements in the development of policy raise additional challenges, not least in defining boundaries of what is deemed ethical by programmers and policymakers. The challenge here is to ensure that regulation maintains a dualistic purpose: pro-innovation and pro-growth, but upholding a basic standard of moral guidance and user security. The innovation-versus-safety dichotomy is not necessarily mutually exclusive. One way to approach this is for regulation to be transparent and coordinated at nascent stages of the development process. Equipping programmers with a clear set of guidelines from the onset substantially reduces the chances of undesirable outcomes. The innovative process can therefore be allowed to develop within the framework rather than be constrained by it (attempts at the latter often occur at an advanced stage of development, when any damage has already been done). The regulatory framework thus needs to ask fundamental questions, such as: What is the purpose of this system? How does it operate and what could be the unintended consequences? How does it integrate a threshold of ethical and safety standards?
Regardless of the exacting regulatory framework, artificial intelligence will be an increasingly potent force in automating and transforming private and public life. AI – and generative AI in particular – will continue to produce the unexpected in solving current problems as well as create new ones. The Judaeo-Christian perspective highlights the centrality of humanity within the rest of creation, and the cosmic order found in Genesis may offer a guiding pathway into the role and scope of AI. A robust approach to AI that challenges its development from the foundational stages, and takes into consideration its aim, nature and purpose in relation to individuals and wider society, leaves us in a more optimistic position to expect positive outcomes in the years ahead.
Notes to Conclusion
[1] Stuart Russell and Peter Norvig, Artificial Intelligence: A Modern Approach, Global Edition, Harlow: Pearson, 2021, p. 1037.
[2] ‘AI Fairness 360’, IBM Research, Trusted AI, 20 April 2023, https://www.ibm.com/opensource/open/projects/ai-fairness-360/.
[3] Adam Thierer, Permissionless Innovation: The Continuing Case for Comprehensive Technological Freedom, Arlington, VA: Mercatus Center at George Mason University, 2014, p. 2.
[4] House of Lords, Select Committee on Artificial Intelligence, AI in the UK: Ready, Willing and Able?, 16 April 2018, p. 7.
[5] House of Lords, AI in the UK, p. 125.
[6] House of Lords, Liaison Committee, AI in the UK: No Room for Complacency, 18 December 2020, p. 11.