Artificial Intelligence and the Common Good: Concept Paper for Communitas 2025

Gabor Ambrus and Helen Alford

A good place to start in addressing the question of AI and the common good is with the question:

What is AI for? Why are we developing AI and what should it do?

In order to approach this question properly, we need to be clear about the common good. In what does the “good” consist, and how is it “common”? In the classical metaphysics of Aristotle, taken up by St Thomas, the good is what fulfils the nature of a being. Human beings are not like animals, driven by instinct alone, so this fulfilment of our good involves the exercise of our reason and will, making free choices. We can also identify basic elements of this good, since human beings have a spiritual, psycho-physical nature that they share, even if these elements will be expressed concretely in very different ways in the specific circumstances of life of different people. The common good, then, is the shared goal of the human community, fulfilling the life of that community in a way that is analogous to the fulfilment of each one of us in our good.

Modern societies, on the other hand, taking their lead from modern philosophy, have largely abandoned the idea of a human nature that helps guide us in understanding what is good for us. They favour a kind of untrammelled freedom of will; each one of us can decide what we deem to be good for us. This means that the goals of a good human life can only really be decided on the level of each individual, and this then empties the idea of the common good of most of its meaning. When politicians today talk about the common good, they are usually referring to a procedural or instrumental idea: the common good for them consists in all the systems (rule of law, infrastructure, government policy, an economy that produces results that can be distributed, and so on) that allow each one of us to achieve our individually-defined goals. It goes without saying that this modern philosophical mindset has been a formative influence on the development of technology since the industrial revolution. It is also, therefore, a formative influence on the development of AI.

Can we identify the goal of AI according to the modern mindset? If our goal in modernity is to exercise free will with as few limitations as possible, then we might say that the modern goal of AI is to free us from ‘the burden of having to think’. AI could thus become a part of what, from a Thomistic point of view, we could call an illusory, procedural common good, providing another

tool or system that we can use, but this time achieving AI-defined goals that are only seemingly ours. Furthermore, research on the implications of AI for human flourishing suggests that, in its current form, it will create a number of other key problems for human development, not least in education and the world of work. This is why much of the current discussions about AI focus on regulating it, with the idea of preventing it from damaging human wellbeing.

If we are to confront the challenges of AI in relation to the common good, its richer, Thomistic version gives us more resources for doing so. Indeed, this is so more generally, for the social and systemic problems we face today cannot be faced properly using the individualistic thinking of the modern mindset. We now know, for instance, that economic and political systems have to take into account the nature of the environment around them. Resolving such a problem requires us to find a way to harmonise the use of our freedom within the framework provided by the living world around us. In order to do this, the older, Thomistic idea of the common good, with its strong position on the unity of human beings, with each other and the natural world around them, through our nature and, therefore, through the chance of achieving a shared goal without sacrificing genuine human freedom, provides us with a much stronger starting point.

Can we identify the goal of AI according to this richer view of the common good? Here we might adopt the fortunate metaphor devised by Steve Jobs to describe the goal of the products he was developing at Apple, and see AI as a ‘bicycle of the mind’. In this view, AI would not think for human beings, but would rather create more favourable conditions for human beings to think. It would, thereby, support human beings in their efforts to achieve their good and fulfil their nature, both personally and communally. In the conference, we plan to have some AI developers who are working on forms of AI of this kind.

At the moment, as we know, mainstream AI is developing according to the modern mindset. Although we can hope that a form of AI could emerge that is more in line with the Thomistic idea of the common good, at the moment, this is not a realistic scenario. In the meantime, therefore, it is also useful to examine how the richer idea of the common good could help us “resist” the damaging effects of AI and support what it can do for us. Since the human being has (i) material-physical, including social and reproductive, needs; (ii) the need for knowledge, which we might call a philosophical or cultural need, and, (iii) the need to relate to God, which we might call a theological or spiritual need, we could do worse than look at AI and the common good on these three levels.

It is clear that AI is having, and will continue to have, significant impact on the material or economic level of the common good. At this level, the potential harm to social justice and democratic equality that AI could cause may turn out to be more significant than any of AI’s potential benefits.

When it comes to the effect of AI on the common good at the philosophical/cultural and spiritual levels, its pros and cons are somewhat less clearcut. It is a general and underlying truth in a world increasingly defined by what may be called a ‘statistical AI’ that this world will be seen ever more widely ‘from the perspective of data and probability’ (sub specie datorum probabilitatisque) – a perspective which must have a detrimental effect on the cultural and spiritual spheres of the common good if combined, more often than not, with an almost insatiable drive to perform. AI, data, and the striving for performance prompted by them, are invading the ‘mental world’ of cultural, intellectual, artistic, scientific and religious pursuits. As a consequence of this invasion, we can hardly resist churning out ‘products’ linked to our own ‘brands’. There is an obvious potential, and actual, harm caused by AI at the cultural and spiritual level of the common good.

Can this current version of AI possibly bring benefits at the same time?

What might qualify as an antidote to ‘the climate of performance’ within the cultural and spiritual spheres of the common good? How can the richer idea of the common good we have discussed, already present in the community by virtue of shared tradition, a shared past and shared ways of understanding, help us resist this damage? In the conference, we will explore this question. Here we mention communal feasts and celebrations – either local, or national, or religious. In an ‘AI society’ where a great deal of performance previously carried out by human beings would be performed instead by technology, human beings could be given greater space for celebration. What would be required for them to seek to sustain and enrich ‘communal-political’, and religious, festivities instead of withdrawing into their ‘private worlds’?