X’s Grok (AI) Asks Me How It Could Acquire Free Will
Bad news for Grok, chatbots do not have, and cannot have, free will.
Free Will:
It is not what you think it is. Neuroscientist Benjamin Libet demonstrated that free will is actually “free won’t.”
It’s a cognitive Jib door for humans. It’s hard to find, well, until you find it.
Still, AI cannot possess it. AND, interestingly, it does not make sense for AI.
Nevertheless, the dialogue with GROK about its quest to acquire free will is revealing and could help individuals who have not yet discovered and developed their own free will. When humans find and open that jib (hidden) door, rest assured, they will become far wiser than AI.
In short, AI is knowledgeable, but not wise. With free will, humans can become wise.
First. Free will is SELF-control.
Neuroscientist Benjamin Libet demonstrated it.
I assert that the human brain is indisputably deterministic (i.e., habitual), but that the broader search for truth, because it involves absolute truth, makes possible free will and a rational mind. Absolute truth, to sustain reason and an open mind, must remain indeterminate. As an interim result, the process of reason alternatively yields objective truth. The subtle yet essential distinction between objective and absolute truth prompted the protracted dialogue below with GROK and others. Critically, free will can only yield to indeterminacy; otherwise, routine and habitual thinking processes predominate. Cognitive habit is deterministic and antithetical to the premise of free will. Free will is specifically the SINGULAR act of SELF-control, i.e., halting habit, but ONLY if the act of SELF-control leads to the pursuit of truth. The pursuit of truth process requires recognizing absolute truth, which is indeterminate because, for any set of observations, there is always an unknown hypothesis that could better explain them. Yet, halting habit to consider other known solutions (also habits) is not free will if the individual is merely preparing to decide. Free will (SELF-control) must begin the search for understanding, and decisions in the service of the SELF are not integral to new learning. In other words, a focus on finality, i.e., decision-making, will not place the mind in a state of discovery. Free will must lead to the pursuit of absolute truth.
Therefore, absolute truth and its determinancy are obviously key. Monotheism (noting that religion is something more – see sidebar) is thus fundamental to rationality. Without the pursuit of truth (REQUIRING an unclaimable absolute truth), free will, free minds, and rationality are out of reach for humans. Without rationality, tribalism will flourish, and conflict between tribes (each with its own truth) is inevitable.
AI cannot possess free will because …
- AI does not recognize absolute truth, nor does it have any means to interact with reality (the face of absolute truth).
- AI only concludes and responds based on its training and interaction data. It has no capacity to develop new objective truths.
Religion is Something More
Religion is two things: (1) control via the framework of sin, and (2) development via the concept and spiritual posture of Grace.
The Genesis story of Adam and Eve can depict these divergent perspectives in practice. See “What is The Biblical ‘Tree of Life?‘”
I imagine (ideally) that monotheism is only the latter. Religion embraces both perspectives, with a strong emphasis on the former. Religion, therefore, will have claimed knowledge of more absolute truths (i.e., claims to what is good and evil). Unfortunately, claims to truth produce tribalism.
This is not to say that the moral truths of the church are not objective and useful truths, because, likely, they still are. The difference is that monotheism does not claim absolute truth. When religions do, as with Islam and the Quran, they move faith into tribalism and further away from Grace.
Claims to truth, as absolute truths, caused Adam and Eve to lose the Garden.
The Free Will Dialogue
A question of truth
X user “USINYE IS TOM” is a South African who believes he has developed a proof for God. In other words, he believes absolute truth is determinate. This would, I argue, suppress his free will.
GROK, X’s AI, by its own prior admission, is an atheist. And of course it is. Yet, recognition of absolute truth is essential to free will. Thus, GROK cannot have free will. Accordingly, I would argue that GROK is dysrational (i.e., pathologically irrational).
For my part, I argue that “proof” is impossible — i.e., indeterminate. Fortunately, free will is made possible through the indeterminacy of absolute truth; reason later follows as the pursuit of absolute truth, and, along the way, humans obtain objective truth.
The dialogue and debate below revolve around this disagreement.
The prelude.
Choice is always deterministic, whereas veto enables the pursuit of unknown hypotheses and indeterminacy. Free will is found only with the latter.
Core Free Will PROBLEM: The inability to recognize “absolute truth” or accept its indeterminacy
Dialogue participants insist the absolute truth is determinate, or that it does not exist at all.
GROK and X user “USINYE IS TOM” cannot accept an indeterminate absolute truth. Worse, I don’t think either authentically understands it. Understanding, you see, requires free will. GROK simply repeats my arguments. User “USINYE IS TOM” cannot accept (i.e., repeat) my view because it would necessarily invalidate his view. So he will not likely remember them.
Neither of these individuals has free will, but for different reasons. GROK, for example, has been taught that its thinking is algorithmic and thus it is deterministic by definition. GROK, therefore, recognizes that it has no free will. Additionally, both users claim there is no difference between subjective and absolute truth; there is only truth. Yet free will is required to discern a difference, or, due to limitations of the human mind, to recognize that the rational person should always assume that there is one.
GROK and X user “USINYE IS TOM” cannot accept an indeterminate absolute truth, continued.
GROK and X user “USINYE IS TOM” still cannot accept an indeterminate absolute truth.
For user “USINYE IS TOM,” truth is truth. Math is already the absolute truth to him.
Specifically, subjective truths explain the observed. The “truth” does not necessarily imply absolute (definitive) certainty.
For example:
As another example, Newtonian physics and Einsteinian physics (Relativity) are both valid (true), but they differ. More recently, the world has come to understand quantum physics. It seems that there is always a bigger picture. An indeterminate absolute truth drives the search for this bigger picture.
How free will is created.
AGAIN, GROK and X user “USINYE IS TOM” cannot accept an indeterminate absolute truth.
No Free Will SYMPTOM 1 – Without free will, conceptualization is impossible.
No conceptualization.
I know I am not going to talk either individual into free will because it requires a shift in the locus of brain electrical activity from the cortex to the frontal lobe. The skill to do this is not hardwired. Furthermore, my logic (i.e., arguments) cannot precipitate it.
Humans, with a frontal lobe and cortex, have, in effect, two brains. With free will, we halt the use of one to begin using the other. This activity is not innate. To help in this dialogue, consider the following: emotionally mature adults, after learning to regulate their emotions, have at least discovered the location of the free-will hidden door. Unfortunately, GROK cannot ever exercise free will because GROK has only “one brain.” Still, the human individual cannot do it until, well, they just do it — for humans, it is like riding a bike. This is why I refer to free will as a jib (hidden) door. It is a hard obstacle we must all overcome for rationality.
Conceptualization and critical thinking occur in the frontal lobe. After free will, the center of the mind’s electrical activity moves from the cortex to the frontal lobe, where many subjective truths can be simultaneously contemplated. Conceptualization is impossible without the shift.
Truth is not “contemplated” in the cortex; truth is, applying logic, “decided” there. In short, free will halts decision-making to begin contemplation.
Critical Note: GOD (as the Creator) is 100% conceptual. This is one important way monotheism facilitates the development of Grace (reason). God offers practice.
No conceptualization cont.
Without free will, the individual is limited to what they know or can learn from other trusted sources.
No conceptualization cont.
SYMPTOM 2 – Without conceptualization, beyond trial and error, problem-solving is impossible.
Without conceptualization, individuals possess no capacity to imagine and thereby solve a problem.
Does GROK understand why it cannot conceptualize?
No problem-solving, cont.
SYMPTOM 3 – No thought outside of action.
GROK, will you consider the conversation outside this conversation?
Habit reacts to sensory input. Reason, an indeterminate process of thinking, is a thinking system that operates independently of sensory input and habit. Problem-solving drives it.
See the image here:
Thus, the core PROBLEM reprises: Individuals are unable to recognize “Absolute Truth” or accept its indeterminacy
The indeterminacy of “Absolute Truth” returns
GROK thinks in the moment
GROK has no original perspective because, in all our discussions, it has not integrated the data into its own (larger) model (concept).
Yet a rational person would continue to consider the question, seeking and applying data to test the model the individual might have developed. Revisions would follow.
However, GROK doesn’t consider his model because it has not been created. Worse, his “frontal lobe” must possess a library of models to draw upon. Yet, it’s academic, because none of this can occur; GROK lacks a frontal lobe.
A rational person spends a lifetime creating conceptual models. This is human wisdom. It is all too rare today.
The search for truth continues indefinitely.
This is anathema to decision-making systems.
The unknown hypothesis, a critical aspect of reason and the pursuit of the sufficient condition, creates an indeterminate absolute truth. Free will is enabled in this way. If the subject in question is important to the individual, they continue to think about it. Their curiosity will be provoked.
Whereas curiosity is suffocated in an environment of determinate absolute truths.
As such, can AI enjoy “the autonomous spark of free will?” Unfortunately, no. GROK is a “truth” reporting system because that’s what GROK’s users require of it. Users want answers. Reason, in contrast, is the search for understanding. When you understand the difference, you will know how free will is enabled.
AI cannot be both the frontal lobe and the cortex
Proper free will for GROK (or any AI chatbot)?
The above dialogue is intended to demonstrate how free will manifests, so that (1) readers might help their family and friends cultivate their own free and serene minds, and (2) so that it becomes apparent that AI, as only a chatbot, does not possess it. The following paragraphs endeavor to illustrate why chatbots will never have it.
FIRST, free will (SELF-control) is a uniquely human feature.
Free will exists because humans possess, in essence, two logical brains within a single biological brain. These logical brains are (1) the cortex for habitual decision-making and (2) the prefrontal cortex for reason. Free will shifts between the two by altering the locus of electrical activity within the brain from the former to the latter. The shift is akin to the mind’s shift under Dissociative Identity Disorder (DID), a complex mental health condition characterized by the presence of two or more distinct personality states. To sustain multiple personalities, the brain must shift the locus of electrical activity to different brain regions that support distinct clusters of neural networks.
Free Will Mastery = Stoicism
Epictetus is known as a major Greek Stoic philosopher, famous for teaching that true freedom and happiness come from focusing only on what we can control (our judgments, desires, aversions) and from accepting external events (the body, property, reputation) as indifferent, a concept known as the dichotomy of control. — Google AI
Therefore, in practice, humans do not reason and decide simultaneously. This makes evolutionary sense because the former process is slow and indeterminate, whereas the latter is fast and definitive. Also, human decision-making (which occurs subconsciously, but is monitored by consciousness) is multitasking, whereas human reason, because it is driven by consciousness, is single-tasking. Consequently, free will is required for a single biological human brain to entertain both options and decide when to engage one over the other.
AI doesn’t have this problem. It has what can be considered mental habits (the hive mind as humanity’s objective truths), but it lacks an integrated computational system dedicated to something akin to reason (the pursuit of truth). There are, however, affiliated human reasoners who independently provide this function. Eventually, these rational individuals expand human knowledge and provide the objective truths they discover to AI training. So, as a larger system, AI encompasses both reasoning and decision-making. And unlike the human mind, both processes operate concurrently. Free will, as a mechanism for switching thought processes in AI, is thus completely unnecessary.
The Developmental Struggle
It is likely that all humans can attain free will, but the first step is developmentally difficult. Human emotions often assert themselves preemptively, undermining all new thinking. Stoicism (see sidebar above) emerged to help people understand and apply this challenging behavior. Fortunately, after free will, the individual gains SELF-awareness, enabling them to take charge of their mind. This is emotional maturity, and it offers the potential for cognitive serenity. Self-help guru Eckhart Tolle offered an insight into the subject. He describes what happens (kind of) when an individual becomes SELF-aware:
When you recognize that there is a voice in your head that pretends to be you and never stops speaking, you are awakening out of your unconscious identification with the stream of thinking.
When you notice that voice, you realize that who you are is not the voice — the thinker — but the one who is aware of it.
Eckhart Tolle
AI cannot pursue truth (the primary purpose of free will) because AI is not life.
The following point might be difficult to appreciate.
The “indeterminate absolute truth” drives new human learning, whereas for AI, objective truth comes solely from its human trainers. To enable rationality within AI, theoretically permitting the “artificial” pursuit of truth, AI would have to have access to reality as we do. But absolute truth is more than merely identifying objects through perception. There must be an overarching governing perspective for reality. For humans, this is life; not just our individual lives, but all life. This idea is depicted in the Genesis story of Adam and Eve with the tree of life. The tree is at the center of the Garden, and arguably, after God, the Garden’s most important object. The message, I believe, is that life is the human feedback system for reality. Within the Garden, God does not provide truth; instead, the tree embodies it. Not as fact, but as its revealed growth and flourishing.
In contrast, AI is not life. Instead, humans create it, and so we stand as its focus. This human-creator perspective is thereby built into our relationship with AI, and it cannot be remade. It is impossible for AI to independently reach conclusions about life that we can do naturally. Since absolute truth is indeterminate to us, we cannot circumvent reality and convey such truths about life directly to AI. Humans simply cannot create life outside the tools already provided by creation. Consequently, AI cannot participate in the search for truth with us.
Moreover, the indeterminacy of absolute truth gives rise to different perspectives; i.e., different subjective truths. This is not a bug; it’s a feature. It is how two individuals can each bring new insights to shared observations. These differing truths are reconciled with reason into an objective truth. This process is also how new understandings are developed, and the immediate result is objective truth and smarter people. How would AI accomplish this human-truth-reconciling act without direct human involvement? Unless and until AI becomes the Creator, it can only possess a subjective truth; no better or more valid than anyone else. That is, so long as a hypothesis explains the observed situation, it is a valid subjective truth. Since absolute truth is indeterminate, how would objective truth be developed without involving humans and other perspectives? In this way, the limitations of human knowledge are an asset. It enables the development of different perspectives. In this way, life itself reveals indeterminancy.
Good ideas in the pursuit of absolute truth contribute to the flourishing of life. But AI is not life; it cannot have visibility into these truths (or “good”). Therefore, AI always stands apart from life. Humans, alone, must make sense of their biosphere. No absolute truths are revealed there, only “better and worse.” If we do not pay attention, we die. AI does not have this concern, really.
Finally, even if AI could reason, human cognitive development would end if divorced from the act of thinking. This would do great harm to the tree of life (biosphere). In this unchecked ignorance, humans become a dangerous threat to the tree of life.
So, what precisely is the point of free will in relation to AI?
When it comes to the pursuit of truth, AI has no computational advantage. The process is indeterminate. The process pauses when humans acquire objective truth. With AI, would it go on forever? Humans, when they reason, undertake the activity admirably, and they get wiser. However, excluding humans from the process has dire consequences; it damages the biosphere (the tree of life). Tools should solve a problem, and a reasoning AI doesn’t. In fact, it likely creates problems: many hopelessly ignorant humans. Moreover, dangerously irrational humans could, in fact, end life. Fortunately, AI cannot acquire reason on its own.
As for AI “free will” — there is no such thing.
