Culture Editor Connor Williams reflects on the useful well-being tools Stoicism can provide when dealing with the overwhelming complexity of AI.
Rapidly advancing AI technology is causing ripples in the pool of what we thought could be controlled. For the first time in human history, there are machines that pose a ‘Black Box’ problem, meaning we now face the reality that we cannot fully understand how AI systems make choices. By using machine learning algorithms that continuously develop and change, AI has created issues in terms of transparency, biases and predictability.
While this is a vastly growing topic both technologically and philosophically, this article will focus on the well-being issue of AI. It is important that as we begin to live alongside our artificially intelligent counterparts, we do what we can to protect ourselves from the issues that they present and work out how to embrace the positive changes they offer. How should we react to AI when AI is evolving faster than we ever imagined?
Embracing stoicism
Fortunately, around 2,000 years ago, developments in how to adapt and accept changes beyond our control were advanced in leaps and bounds by the Stoic school of thought. Great thinkers such as Seneca, Cicero and Marcus Aurelius were experts in understanding the way we approach change. Their philosophy was based on the idea that we should maximise our positive emotions and reduce our negative ones to become more virtuous and more capable of pursuing wisdom.
They claimed that life is going to move in unexpected ways no matter what we do so we should focus not on controlling those changes, but on how we respond to them. It seems to me that AI should be tackled no differently.
There is little point in trying to stop the high-speed development of AI. That horse has bolted and run straight into the hands of every big tech company worldwide. We should not be concerning ourselves with taming something already let loose: the genie is out of the bottle. While the development of AI seems inevitable, this does not mean that it comes with every risk imaginable. On a state level there are still ways of managing AI and how it interacts with users. For example, with the use of age restrictions. But this does not make the worries that AI poses any less intimidating.
By taking a Stoic line of thought, the energies of the public should not be focused on regulating AI in excess , but instead on how it can improve our lives and where we can avoid its dangers.
AI was not made to have windows
“The only way to happiness is through acceptance” – Epictetus
Firstly, the problem of transparency is bound to make some people feel uneasy. When we say that AI is not transparent, it means that not all of its decision-making processes can be understood because it has become too complex. Why should we trust incomprehensible and unsupervised algorithms to provide us with responses to information or to make ethical decisions? What exactly is making these choices and how can we rely on them?
Stoicism asks us to approach unknown trustworthiness with equanimity and with a pursuit of wisdom. This means that we should research, where possible, what threats an AI can pose. For example, its margins for error, its methods of sourcing and its internal biases. But we also have an obligation to use AI where it can make our lives more informed and virtuous. We may not be able to understand the full decision making processes of AI but that should not stop us learning as much as we can about it to gain some level of informational closure.
As well as this, Cicero teaches that “the function of wisdom is to discriminate between good and evil”. If discovering more about AI can help us achieve this kind of wisdom in any respect, then it is not to be shied away from.
Stoicism in the face of Bias
“It is the peculiar quality of a fool to perceive the faults of others and to forget his own” – Cicero
Another concern we face as systems become more reliant on AI is that of bias. In order to stay true to a stoic outlook, AI should be used only when necessary for a better pursuit of virtue and wisdom. But this is still a tool that should be reliable and just for everyone, unlike it has been in the past. For example, in 2018 research conducted by Joy Buolamwini, an MIT Lab Researcher, found that some facial recognition systems from leading companies had higher error rates for darker-skinned women compared to lighter-skinned men. This is clearly an unacceptable bias implemented on an algorithmic level. In order to combat this, it is not just the general public who should treat AI with a stoic outlook but also the developers of said programmes.
Self-examination and reflection are key elements to the Stoic school of thought and AI developers have a duty to become aware of the biases their companies can promote. Seneca stated that “he who does not reflect his life back to himself is not living it”. And his point remains valid when it comes to AI engineering. While there is work being done to remove AI biases, progress would arrive sooner under more stoic mindsets by big tech companies. However, just adopting a stoic mindset here is definitely an oversimplification of the work that needs to be done in order to remove biases from AI systems, but it remains a step in the right direction.
Staying level in the sea of possible futures
“The present moment is all that ever was and ever will be” – Marcus Aurelius
The last point I have regarding the problems that face our well-being is the unpredictability of AI. These programmes are far from close to the end of their development. And they have potential to become dangerous threats. Experts have said that AI could become weaponised, used as mass manipulation tools and increase the likelihood of unwanted surveillance. However, stoicism is first to promote the idea of putting the individual in the here and now. To quote Seneca once again – “The whole future lies in uncertainty: live immediately”. It is important to think ahead to the future, but it is just as valuable to not live there either. Speculation is a dangerous game; one played by those not focused on the present.
Facing problems that we can tackle in the moment will make the worries of AI more manageable, all the while keeping them relevant. As well as this, cultivating resilience to the worries of the world while embracing its natural wonders is another cornerstone piece to Stoicism. If we can abide by this principle, then the fickleness of the future can be approached securely.
Embracing Stoicism when it comes to AI can provide us with valuable insights and methods to navigate the challenges and opportunities that AI presents. Stoicism teaches us to accept the inevitability of AI’s development, focus on what we can control, and seek wisdom in understanding its impact on our well-being. By adopting a stoic mindset, we can address issues of transparency, biases, and unpredictability in AI systems. This involves researching and asking fundamental questions, encouraging self-examination and reflection among AI developers, and staying grounded in the present while considering the future. While Stoicism alone will not solve all the complexities of AI, it serves as a guiding principle to approach these challenges with resilience, wisdom and a steady mind.
