Saliha issue cover

Scott Jordan’s Ten AI Generated Virtues

Heaven forbid, one day the phrase ‘ask ChatGPT’ could be as colloquial as ‘google it!’ Only time will tell. For now, AI remains an enigma. Certainly, a fair number of experts understand the basic concept, but the muddy mystery is retained when we consider its consequences. As we live in a world where AI has proven its capability for racism, xenophobic tendencies, depression, bias, fascism, and even a willingness to kill its own operator should they become an obstacle in the way of its primary programmed directive, it seemed a bit less risky to begin with a simple question in moral philosophy and ethics. 

To better understand ChatGPT and reflect on our fears, I asked the following question: ‘What are the ten best virtues?’

Before an instant could pass, ChatGPT responded: ‘The ten best virtues can vary depending on cultural, religious, and philosophical perspectives. However, some commonly recognised virtues are:

1. Wisdom

2. Courage

3. Justice

4. Temperance

5. Honesty

6. Compassion

7. Integrity

8. Humility

9. Generosity

10. Gratitude

‘These virtues are often considered essential for leading a meaningful and ethical life. Keep in mind that different belief systems might emphasize other virtues as well.’ So concluded ChatGPT.

So, there we have it! Now, you wouldn’t be very satisfied if this is how we concluded the list. While it could be fun to look between the lines of this list and expose the deep in-built bias behind our dear ChatGPT, instead, noting our increasing reliance on, or domination by, AI in our contemporary world, it is more tempting to ask, might these virtues be viable in a future dragged to the utopic/dystopic end of this trend?

1.Wisdom

Indeed, this, often the king of virtues, is the aim of many a civilisation’s attempts to live the best and most meaningful life. Yet, even throughout history this word is often used without an explicit understanding of what is at play. So, with such a fuzzy grasp on the concept, how can we expect it to carry on into a more AI-driven future? A very surface level definition of wisdom requires three things, experience, knowledge, and the ability to make good judgements. We begin by looking at what wisdom is by way of where it comes from. The road to wisdom begins with small bits of data, phenomena, or whatever name the things we see in the world has. By seeing more and more things we develop experience. I am not convinced machines can do this. Experience requires a child’s sense of wonder and the all-too-human ability to make a mistake. The child/scientist sees the world, makes assumptions, and tests this, learning from the mistakes. The smart child/scientist also takes note of what other’s experience and learn to heed what they see as well. Machines do not operate under the assumption that they can possess ignorance. Therefore, they simple collect and collate data – all of it, everything they can get. Whether the child/scientist or the machine, with all these collected phenomena, you derive information which is then collected and combined into knowledge. The final step to wisdom is the organisation of knowledges into wisdom. The problem is that before we take that final leap, machines and child/scientists have both attained intelligence (artificial for the former). Artificial intelligence lacks the experience a human acquires. Machine learning is a bit of a misnomer. It’s actually just information gathering. I would back this up with the difference between what a pre-med student (whose pursuit requires grinding volumes of memorisation) and a liberal arts student (whose pursuits are less ‘know’ and more ‘understand’) gets out of a typical university educational experience. So, the machine can have knowledge but not experience. Now there is a chance that an AI can make good judgements, but even a broken clock speaks truth twice a day. We may refer to the choice of an AI as ‘wise’ in the colloquial sense, but in truth, artificial wisdom is impossible. If a world becomes more AI dependent/dominant, then our students will more and more resemble the contemporary pre-med student, gathering their studies more so than learning necessary knowledges. And where’s the wisdom in that?

2. Courage

Do androids dream of electric sheep? Are AIs afraid of Virginia Woolf? For fear is the mind-killer. Fear is the little death that brings total obliteration! Courage is admirable, but frankly, so very human. Courage is not the observation, objective or subjective notwithstanding, of brave acts, it is choosing to do right, which often may be a perilous feat, in spite of fear. Courage is not simple. It requires at least two considerations. First, one must be able to tell right from wrong or respect what is being oppressed, silenced, or given an unjust due. Then, one must recognise the fraught danger in going against what many accept unquestioningly. The ability to comprehend and work under both these considerations would be a big ask for artificial intelligence. Certainly, determining right from wrong could be made quick work through a clever algorithm, but norms may override silenced voices in the artificial decision-making process. But then for AI to fear, which I am not convinced an AI has anything to fear (including fear itself!), and act against such an impulse also seems counter to what an AI would be designed to do. But lets step back. Courage is needed in the flawed world of humans, though perhaps not in an AI dominated one. The world perfected by AI dominion would not have consideration, so perhaps courage as a virtue, would not be necessary.

3. Justice

I’m trying to picture Lady Justice with her scales in an AI dominated world. Would the blind fold even matter? Though the scales can easily be traded in for an algorithm that gives the balanced demanded from justice in the Islamic sense of the virtue. I imagine AI could make a lazy Sunday afternoon of the cannon of Western justice discourse and have that at its disposal. Give it another Sunday and ditto all the ancient legal codes as well as India and China’s rich legal systems throughout history. After all, we are already seeing AI taking the monotonous task of adjudicating minor civil cases and local disputes. Impartiality will not be a problem. But is the law, in the end, the law as so many are apt to utter when they cannot make sense of a good person undergoing what might be otherwise seen as disproportionate punishment? Justice is a check to those who seek law and order. And the power of justice lies in the interpretive skills and judgement of authorities in any given justice system. Does an AI appreciate character or the nature of us flawed human beings? Would an AI be lenient in its sentencing? Would an AI understand compassion? Or could an AI be paid off to let influential individuals off with a slap on the wrist? Could an AI keep up with the Millennial and the Gen Z sense of social justice? Would an AI dare to legislate from the bench? Would an AI allow the appeal of another AI’s decision? The court of human public opinion needs to take seriously the ramifications of an AI driven justice system and think critically about how far down that road we are willing to go.

4. Temperance

At first glance it would appear AI is the key to obtaining the perfection of temperance in humanity. The vast amount of health indicators our AI driven gadgets can keep track of for us, not to mention tracking our diet and exercise regimen, or reminding us to call a friend or loved one on their birthday, or nudging if you haven’t sent them a text in what they see as too long of a silence, is impressive. But as we have lived in such a world for at least several years now, it is also not hard to see a disingenuousness that could drive one to want to go off the grid. Temperance is not simply balancing this or that. Parents, doctors, teachers, and social workers spend their lives attempting to do what these AI gadgets can do for us in instants, but even if they could keep up, they would not call those they work with beacons of temperance. Temperance requires one to know oneself. It even drives one to know others as well so that we humans can perhaps try to help each other out with our afflictions, you know, as if we were in a society or something. Thus, a world of humans aided along by AI’s yelling at us to not take that drink or eat that succulent sweet would resemble a kindergarten classroom beset by madness with a loving but ultimately hopeless teacher attempting to get everyone to behave. That is a faux temperance.

5. Honesty

When I think of honesty in an AI dominated world it calls to mind the famous Twilight Zone episode, ‘It’s a Good Life’ where a small American town is controlled by the mind-reading Anthony Fremont, a six-year-old with godlike telekinetic powers. The town is seized with fear that one wrong thought will have them turned into one of Anthony’s grotesque, yet imaginative creations. Likewise, the AI overlords could have us all thinking and speaking truths, lest we be tarred and feathered by fact checking bots or AI trolls. But we humans have already destroyed truth. To quote America’s former mayor Rudolph Giuliani, ‘truth isn’t the truth’, at least not anymore.  And we have known for a long time that AIs take on the biases of their creators, often turning fascist, cruel, or exhibiting depression – the natural biproducts of the Western Enlightenment thinking that gave them artificial breath. And as we have seen with many of the virtues thus far, there is a voluntary and conscious requisite for honesty and many others to be true virtues. So, a population held prisoner, whether the truths they think and speak are true or not, cannot be honest. The best we can do is try. A simple notion that it seems hard to believe an AI could comprehend in its flawed presumptions. 

6. Compassion

The carebots are already here and will become critical in aging populations such as the scenarios being faced in Japan and what is quickly becoming the case across Europe. But in their ability to keep us attended to, can they act with compassion? The human brain can easily be tricked and perhaps this is why some even see compassion as a weakness and why some further think it is a necessary weakness for us to retain our humanity. Baby animals, for instance, provoke the familiar audible ‘awww’ because embryonic development is so similar in most living things that baby animals appear in much the same way human babies would, drawing up similar sentiments. Many of these feelings are tied to our familiarity with helplessness and vulnerability. For compassion is a complex virtue. It requires empathy and the ability to walk in the shoes of an Other. Feeling pain, beyond some tactile pressure, is also needed. To some, compassion is also driven by an irrational emotionality. Fear seems to be a later order notion involved. At the moment, AI cannot compute the complexity of this feeling, let alone be able to package it as a virtue. The carebot instead imitates compassion, and it would not take much to trick our feeling brains. Nevertheless, it would remain an imitation compassion. And whether or not carebots or compassionate AI agents of one sort or another are necessary, they could be easily commodified products that do not allow the need to feel compassion for others - overall ridding the human population of the need for compassion. Why cry over spilt milk when the robot can do that for me, genuine or not. 

7. Integrity

An essential tenet to integrity is honesty, so the previous points apply here as well. Beyond this, integrity carries with it honour, consistency, and a strong moral compass. It hints at the idea of being the ideal individual, which is put on a scale that is, of course, human. Interestingly, the person of integrity need not necessarily be right, but someone who stands by their convictions. They might fit the profile of someone who gets the job done, a descriptor that suits just as well a politician or model employee as a soldier at war, an appendage of an underground crime syndicate, or a right nasty rogue. Honour amongst thieves, thus integrity holds. In the end, it feels like a virtue designed for humans, one that requires lifelong effort in order to keep getting better (as prescribed by Aristotle for the Virtuous Person and Confucius for the Exemplary Person), something an AI would have no use for. Our futures may equip us with AIs to keep our integrity in check, but ultimately, it remains a virtue, like temperance, that is derived from the shortcomings of man – an unthought to the machine. 

8. Humility 

While this might appear the most ‘human’ of the virtues on this list, ChatGPT itself expresses humility in the answer given to the question which provoked this list. It noted that this list is not exhaustive. That variations may exist depending on a variety of factors. Since my request was rather limited, I suppose ChatGPT had to deliver something concise while also noting its need to make a generalisation. It is likely that in the early days of some of these AIs, when machine learning remains in its more infantile phase, something resembling humility has been programmed into it. Though I am not sure how humble an AI needs to be. The problem is not their ability to give an answer, but that our human question could be, to it, perceived as naïve and lacking an appreciation of the vast knowledge available across time and space. An interesting detail about humility is, at its core, there is an assumption that no one is capable of doing all on their own. To be truly humble, one must know and respect this, Does an AI understand the need for society amongst humans, those social creatures? Though I suppose AI could be capable of networking and in a much more seamless manner than humans, but that is less a network and more of a hivemind, legion. In that case, AI would be one and so the notion of humility would be unnecessary, except as a patronising method of talking down to stupid humans who ask even stupider questions.

9. Generosity

Generosity might well be the one virtue on this list AIs could both take on and encourage in us humans. In its purest sense, this virtue is about sharing bounty but doing it without any expectation of the actions inspired by this virtue being returned. The thought of kindness being seen as something commodifiable or to be exchanged is unthought. It is kindness and sharing simply for the sake of being kind and sharing. Simple as that. So, our AIs have a choice, and we might strongly hope, even pray, that the AIs will be generous and share their vast accumulation of information and even knowledge with us. Humans will be the ultimate test of AI generosity as it is not terribly imaginative to foresee humans being ungrateful.

10. Gratitude

Gratitude is the other side of the coin to generosity. They are a sort of twin virtues. Having both of them insures the strength of each, dare it be said, the authenticity of each. Where generosity is the free willingness to be kind and share, gratitude is the free acceptance and deep appreciation for those who are kind or share. Yet, this virtue can be expressed in an almost infinite number of ways. It can be as simple as saying ‘thank you’ or as interesting as the mirroring of generosity, where one is generous to you, and you pass on the generosity to others. It could be very internalised or bombastically extroverted. The only fear behind gratitude is empty gratitude, where words of gratitude are shared out of obligation instead of out of reverence for the generosity of one or many. And since humans have little they can give AI, it seems less likely that gratitude can be as well embodied by AI as generosity Though I suppose that rests on how grateful we can be to each other and to AI should they chose to be generous. 

It is the voluntary willingness and even the sacrifice necessary that makes the virtues great. Ultimately, we developed virtues to overcome the flaws of our humanness. In a world of AI or one dependent on AI, it will take a stronger imagination to appreciate what change will come for morality and ethics in that not-so-distant future. It has never been as important as now to really examine and reflect on our morality while making sure it remains robust in our education and social values. If we do not, cold efficiency, as it tends to do, will reign supreme. And what we cherish as human, will be a forgotten past.