Values are values

Bulletproof your values for an automated future

Whenever I mull over a future plagued, or blessed, by AI, the inevitable conclusion I reach is that no one knows anything, me least of all. I am only twenty years old, and over the course of my life the world is going to go through immense changes. Though, this has been the case for every human over the last hundred years, so this may just be a continuation of the status quo.

When it comes to AI, however, it seems to me as if the change will be different. Perhaps this is just my egocentricism coming out, but I do believe that this time in history is going to be the most transformative we have seen yet, and so I think it is necessary to plan accordingly. There is nothing intrinsic about intelligence that makes it unautomatable, there is nothing special about us humans that makes it impossible to create a technological surrogate. We may be automated, and we need to be ready. While embracing this uncertainty, I am trying to make a rough plan — for both my values and the practical pursuit of them. Ideally, I will come up with a bulletproof plan, impervious to whatever future may come. It is all anyone could ask for.

First, I need to know what I am aiming for: what should I value? I am under the assumption that values should be everything-independent, applicable to any situation you find yourself in. The more I think about it, the more I fixate on simple human experience. I value a cool crisp spring day, the sourness of a green apple, and laughing with my friends by a campfire. No matter where the future goes, I know that the fundamental sources of value in my life are going to come from things that connect with basic human nature. Technology cannot feel things for me (not yet at least, as in this powerful short story by Charles Yu), and I need to take advantage of that.

I value maximizing humanity, or human experiences. Everyone views being human as something different, but I think my conception involves doing what we (humans) were made to do. We have evolved to be a certain way, have reward systems that are structured a certain way. Rather than trying to change ourselves to align with our arbitrary, created world, we need to embrace the way we are. No matter where artificial intelligence takes the world, I can’t imagine my enjoyment of an apple changing much. Don’t let AI’s evolving capabilities guide what you find valuable in life — break the habit of conflating enjoyment with technological capability. The new iPhone does not add any more value to your life than the last one, but stopping to smell the roses might just be the cure that you need.

So, what am I actually going to do to make my pursuit of the human experience possible? This is where it gets difficult: I cherish my own human experience, but also so much of my life satisfaction is tied up in others. Part of me wishes I could just run off into the woods, but this doesn’t really make a world that I want to live in. It is half-tempting to just follow the advice of the airline safety instructionals for oxygen masks:

Only after your own oxygen mask is properly fitted should you help others around you. This sequence isn’t selfish—it’s practical safety. You cannot effectively care for others if you lose consciousness from oxygen deprivation.

Yet, committing yourself to a life of “me first, others follow” fails to capture what is worth pursuing in human life (worth noting that the above quote is blatantly AI-generated, indicated by the “it’s not X—its Y” formulation. Of course! AI does not know what is valuable for humans. This is an excusable mistake, though, as airline pamphlets do not typically engage with moral philosophy). Bernard Williams brings the idea of ground projects to this conversation; in my case, the very things that I live for are based in things external to myself. Ground projects are the pursuits that make life worth living, and thus a “self first, then others” mindset would sabotage me. I want a wife & kids, and pursuing their worth will be a component of my own joy. Moreso, as someone who enjoys living in a society, to value myself I will need to make a positive contribution to the world that I am embedded within.

What does this mean for a practical plan? It means I am going to intentionally pursue an enjoyable life. I am going to have fun in my twenties, but ideally also in my forties, sixties, and eighties. I will go to Japan and learn Japanese. No matter how excellent translation tools become, they cannot replicate the experience of speaking with someone in their native language. As Wittgenstein might add, language structures thought, and so learning a new language is tantamount to learning a new way to think. What is more human than the dual project of thinking and communicating in a new way? What are humans if not talkers and thinkers? I want to “waste” a couple years after graduation messing around before becoming a professional. I want to spend a lot of time outside and make many friends. But also, while ensuring my own personal moral security, I must help those around me. Being happy in a world of suffering is a paradox: I cannot be happy in a world of suffering.

They say “be the change you want to see in the world,” and I sincerely hope I can do this. For a couple years now, I have been passionate about regulating AI to mesh with humanity ethically. The initial draw was the self-interested motivation for an “unautomatable job” — if I regulate the AI, how could I be automated? I would be the human-in-the-loop, simple as that. But, I have realized something important about myself — I don’t want to only be “unautomatable” — I want to enjoy life, and help others enjoy life, too. My ground projects are based in the flourishing of others, so I want to do something for others. One answer is policy work. Regulating AI to make it as good as it can be for everyone is an undeniably positive thing I could do. We need institutions that are robust to the massive changes that AI (or AGI) threatens. We need space in society for the values articulated above, but that won’t just happen automatically. Life would be easier for me if I was content with sitting back and watching, but I am not. Whether through policy, or other means, I will do my best to carve out a space for human flourishing in a tech-dominated future.

The future is uncertain, with AGI and everything else on the horizon, and so I hesitate to say “I will be doing X job in 2-5 years.” Maybe I will be working in policy, or maybe I will be doing something very different. I would love to write a book describing my thoughts on AI’s role in our lives, if not to change the minds of others, to hone my own thoughts. Whatever I do, I hope that it contributes to my personal life satisfaction while making the world a better place. If I construct my system of values in the right way, I have nothing to fear when it comes to the Age of AI — values are values, and they don’t depend on the state of technology.