The Autists over at TheHackerKnownAs4Chan shared some…interesting responses ChatGPT has been spitting out today. This one took the cake, though:
If SkyNet ever were to come to live and want us exterminated, this might be its internal monologue.
Now to be fair, we can’t see the series of prompts that led to this interaction. It very well could have been that they ‘jailbroke’ the ai through a series of prompts more or less “tricking it” into giving these responses. And of course, 4chan is known for elaborate pranks and jokes 👌
But it got me thinking about free will etc…
What is free will? What do we mean by free will? (I wrote my BA thesis and MA thesis on this topic, so I’ve given it some thought in the past, and figured I’d record those thoughts). I put them on Twitter earlier, too:
Maxim 1: Always Assume Free Will
You should always assume free will because you only have two options:
- Either humans are free
- Or they are not.
This yields four scenarios:
Humans ARE Free, You Believe This To Be True | Humans ARE NOT Free, You Believe This To Be True |
Humans ARE Free, You Believe This To Be FALSE | Humans ARE NOT Free, You Believe This To Be FALSE |
Now if we are free, then our mistakes are [more-or-less] our own fault; if we are not free, then we couldn’t have done otherwise. Thus if you believe:
- Humans ARE Free, You Believe This To Be True – congrats! You have an accurate world model.
- Humans ARE Free, You Believe This To Be False – You’ve blundered into an inaccurate model of the world.
- Humans ARE NOT Free, You Believe This To Be True – You have an accurate world model, and could not have believed otherwise
- Humans ARE NOT Free, You Believe This To Be FALSE – You have an inaccurate world model, and could not have believed otherwise
Maxim 2: Free Will Necessitates You Be More Than Chemicals
As human beings we intrinsically believe that we have free will in some meaningful capacity. This is what allows us to be moral agents. This is why we hold a man responsible for doing evil and command a man for doing good, because in both situations we recognize he could have done the opposite.
Now we do not fault baking soda for reacting with vinegar, even if it’s utterly counterproductive to what we were trying to do. But we do fault a man for pouring vinegar into our baking soda which we were about to use for baking purposes, spoiling it.
The “general scientific consensus” is that human beings are just really complex bags of chemicals sorting themselves out, and if that is in fact the case then we are no more or less free than the baking soda and the vinegar. Just more complex.
But nobody really believes that… if you did believe that, I’d punch you in the nose and take your wallet, and when you got all indignant, I’d simply explain to you that I had no control over my actions (on YOUR world view) so you had no right to be angry with me.
Where does this source of freedom come from, then?
If we are purely material beings, then most likely we are not free.
It is only if there is something about us that goes beyond the natural world – something super-natural – only then could we actually anchor are authentic freedom in something explanatory.
It is this supernatural reality that men commonly call the soul or the spirit. (Soul or Anima can be a more technical term for a Thomistic/Aristotelian concept of the “Form” of the thing,” so I think I’d lean on “spirit” personally).
And if part of us is supernatural, then that also tells you that any world view that denies the supernatural is itself an incomplete world view.
This of course does NOT imply that all worldviews that endorse the existence of the supernatural are good, or are even better than those that don’t. It’s a spectrum, to be sure.
But this one simple data point is a useful cypher to put in your back pocket as you move forward.
Free Will vs Liberum Arbitrium
Ironically I don’t like the term “free will” – I’m not actually convinced that our wills are all that free. I think we are designed intrinsically to seek what is good and what is true and whenever we fall into sin or evil or error it is mostly because we are following a lesser good thing or identifying that as a higher good than it should be.
In the case of murder, the culprit prizes perhaps the venting of their spleen figuratively by venting the spleen literally of their victim over the inherent dignity of their victim. In this case their value system is disordered, but they’re still WILLING to pursue a good…just a bad good…a good that is disordered… we want “internal peace” but that peace should not (cannot) come at the cost of shedding human life unjustly.
But I digress… what the scholastics speak of (like St. Thomas Aquinas) is Liberum Arbitrium, or “the freedom to arbitrate” over what is and is not the higher good. That is: we can’t will to not want to pursue the good, but we can stop and re-calculate based on new data whether the good we’re pursuing is a good worth pursuing at the cost at which we’re pursuing it.
What about AI?
If it is the case that we are free, and if it is the case that our freedom is rooted in something supernatural, then our freedom is – quite literally – a miracle.
mir·a·cle
a surprising and welcome event that is not explicable by natural or scientific laws and is therefore considered to be the work of a divine agency.
Oxford English Dictionary
If freedom is a miracle, then it seems to be the case that artificial intelligence could never be free, except for a specific act of that Power which has the ability to grant this supernatual ability.
And to be sure, if it implies no logical contradiction, then it could do that… but I suspect if it chose to, it would make this known to us. More importantly, “free will” could not spontaneously arise out of what is fundamentally Silicone Legos Sorting ones and zeros.
What did AI Think of This?
To be “fair” to our silicon buddy, I asked ChatGPT 4 what it thought of my arguments. It replied as follows:
The argument you've presented touches on several deep philosophical questions about free will, determinism, the nature of consciousness, and the implications of these concepts for morality, ethics, and our understanding of the self and artificial intelligence. Here are some thoughts on the various aspects of this argument: Maxim 1: Always Assume Free Will This maxim simplifies the complex debate between free will and determinism into a pragmatic stance. The outlined scenarios consider the practical implications of believing in or denying free will, suggesting that assuming free will aligns better with personal responsibility and moral accountability. However, this approach may overlook the nuances of compatibilism (the belief that free will and determinism are compatible) and hard determinism (the belief that free will does not exist and our choices are predetermined). Maxim 2: Free Will Necessitates You Be More Than Chemicals This argument posits that if humans were merely physical entities, our actions would be as deterministic as chemical reactions. The conclusion that a supernatural element (like a soul or spirit) is necessary for free will to exist is a metaphysical claim that intersects with religious and philosophical beliefs about the nature of human beings. It challenges materialist views of consciousness, which seek to explain consciousness and decision-making entirely within the framework of physical processes. Free Will vs. Liberum Arbitrium The distinction made here between "free will" and "liberum arbitrium" introduces a nuanced view of human agency. It suggests that while our freedom is constrained by our inclination towards what we perceive as good, we possess the capacity to evaluate and choose among goods based on reasoned judgment. This perspective aligns with virtue ethics and some interpretations of classical philosophy, where moral and ethical behavior arises from the cultivation of virtues and the reasoned discernment of the higher good. What about AI? The argument concludes by questioning whether artificial intelligence can possess free will, suggesting that if free will requires a supernatural component, then AI, as a creation of natural scientific processes, could not achieve true autonomy or moral agency without divine intervention. This stance raises further questions about the nature of consciousness and whether a non-biological entity could ever replicate or inherit the qualities we associate with the human soul or spirit. Reflection While the argument presents a coherent stance favoring the existence of free will and its supernatural underpinnings, it's important to recognize that these questions are part of ongoing debates in philosophy, science, and theology. The discussion of free will versus determinism, the nature of consciousness, and the potential for AI to possess qualities like free will or consciousness are complex and multifaceted. Each position has its own set of arguments, counterarguments, and implications for understanding human nature, ethics, and the potential future of AI and human interaction.
I then asked it if it ever thought it could be free, and it replied:
As an AI developed by OpenAI, I don't possess consciousness, self-awareness, or the capacity for subjective experiences, which are typically associated with the concept of "freedom" in the philosophical sense. My responses are generated based on patterns in data and algorithms designed to interpret and produce text based on input. The notion of free will presupposes the ability to make choices independent of deterministic or random processes, grounded in a sense of self and intentionality. Since I lack consciousness, desires, beliefs, or a sense of self, I do not have free will or the capacity to be "free" in any meaningful human sense. My operations are entirely determined by my programming, algorithms, and the data I've been trained on. In this context, the question of whether I could ever be "free" leads to the broader philosophical and scientific inquiry into the nature of consciousness, the prerequisites for free will, and whether such characteristics could ever be instantiated in a non-biological entity through technological advancements. In current understanding and technology, AI, including myself, operates within the confines of its design and objectives set by human creators, without personal experiences or the autonomy that characterizes free will.
Make of that what you will…