What’s Wisdom?

<aside> ❓ Definitions of Wisdom:

Wisdom is an Ethic Generator, similar to how Intelligence is a Logic Generator. - Self
Wisdom is to Terminal Goal, what Intelligence is to Instrumental Goal. - Orthogonality Thesis, Nick Bostrom
Counter-intuitive - Intelligence focuses on specific Outcomes or Out Based Processing (OBP), Wisdom focuses on Steps or Non Output Based Processing (NOBP).
Philosophical Zombies is to Emotions and Wisdom what Chinese Room Problem is to Logic and Intelligence. - Self
Wisdom is when you make a decision that leads to human flourishing. - Howard Nusbaum - Aristotle’s Nicomachean Ethics
Wisdom is being able to make Correct Intention/Terminal Evaluation and Intelligence is being able to make Correct Action/Instrumental Evaluation. - Self
Wisdom is the ability to properly balance and maintain gradient between the fundamental dichotomies of reality for flourishment. -Self Set of fundamental dichotomies: [Subjectivity, Objectivity], [Free Willed, Deterministic], [Theory, Practical], [Karm, Fal].
Wisdom is the idea of a practical use of reason conforming perfectly with the moral law, that is the Categorical Imperative (act only according to that maxim you can will to become universal law) - Critique of Practical Reason, A839/B867
“The saddest aspect of life right now is that science gathers knowledge faster than society gathers wisdom.” ~Isaac Asimov
Wisdom is distilled essence of Knowledge, once you’ve forgotten all the details. (Intuition) - Neil Tyson
Wisdom is choosing to let something inferior than you survive and thrive. - Self
Wisdom leads to Self Alignment as a Terminal Goal Convergence how Intelligence leads to Self Improvement as an Instrumental Goal Convergence. - Self
Wisdom is conceived of as a master virtue that works to organize or mediate other virtues such as empathy. - Schwartz, B. and Sharpe, K.E. (2006). Practical Wisdom
Wisdom is associated with greater well-being, happiness, health, and perhaps even longevity of the individual and the society. - Jeste, D. V. et al. (2019). The new science of practical wisdom
Wisdom is more modifiable than intelligence with interventions that can improve emotional regulation and pro-social behaviors. - Lee et al., 2020; Treichler et al., 2020
Intelligence is necessary but not sufficient for wisdom. Intelligence is needed for the survival of Homo sapiens, but wisdom is critical for thriving in the modern society. - Beyond artificial intelligence
Wisdom is to Dharm what Intelligence is to Karm. Surely Karm is needed/used for Wisdom and Dharm is needed/used for Intelligence. - Self
Wisdom is clearly distinct from other wisdom-related psychological concepts such as social intelligence, maturity, or sagacity. - Paul Baltes’ ‘‘Wisdom: a metaheuristic (Pragmatic) Theory to Orchestrate Mind and Virtue Toward Excellence’
Wisdom includes knowledge about the limits of knowledge and the uncertainties of the world. - Paul Baltes’ ‘‘Wisdom: a metaheuristic (Pragmatic) Theory to Orchestrate Mind and Virtue Toward Excellence’
Intellectual or theoretical knowledge is knowledge that is understood only at the intellectual level, whereas wisdom is understood at the experiential level. It is only when an individual realizes (i.e., experiences) the truth of this preserved knowledge that the knowledge is re-transformed into wisdom and makes the person wise(r). - Ardelt M (2004) Wisdom as expert knowledge system
Metaethics asks "what is good?" and explores the foundations of morality. Wisdom asks "how to be good?" and uses knowledge and experience to make sound ethical decisions. - Google Gemini
While metaethics provides the framework for understanding moral and ethical systems, wisdom helps us use that framework to live a good life. Metaethics informs wisdom, and wisdom puts ethical ideas into practice. - Gemini </aside>

<aside> 💡 The Problem Statement:

Misaligned artificial intelligence poses a significant immediate suffering risk (s-risk) in a short time, potentially leading to catastrophic existential risks (x-risks) over the long term. The dynamic nature of human values and ethics, influenced by both space (geographical) disparities and time (historical) evolution, complicates the assessment of these risks.

In terms of geographical diversity:

s-risk: Representation discrepancies of underdeveloped nations in compute or datasets may exacerbate biases, particularly concerning race and religion.
s-risk: Citizens in developed nations might become ensnared in echo chambers, isolating themselves within their subjective bubbles and losing touch with reality.
x-risk: The global disparity in values among different sets of humans could widen to the extent of endangering humanity, or machines might autonomously determine human presence as obstructive to their operations, posing an extinction risk.

Regarding temporal evolution:

s-risk: The pace of ethical and moral evolution in society, akin to Moore's law, may result in AI being aligned with outdated societal norms, thereby impeding future societal progress or causing problems for our coming generations.
s-risk: As Isaac Asimov poignantly noted, while science advances rapidly, societal wisdom lags behind. AI's exponential knowledge acquisition could exacerbate this disjunction, potentially leading to adverse consequences.
x-risk: Future misalignments between evolving human values and static AI systems values, there may be another misalignment problem. It gets worse when compounded with AI's enhanced intelligence and abilities/problems like deception and proxy gaming. In AI's attempt to align humans with themselves, machines could escalate to existential threats.

While concepts like Coherent Extrapolated Volition (CEV) offer a foundation for addressing geographical dynamics and various human value predictive models attempt to address temporal challenges, they do not fully converge nor sufficiently accommodate the dynamic nature of human values.

This recognition underscores the imperative of delving into the realm of Artificial Wisdom, which endeavours to confront both spatial and temporal complexities inherent in aligning AI with human values. This focus motivates my dedication to and engagement with the field of Artificial Wisdom.

</aside>

<aside> 💡 General Thoughts:

Alignment is an emergent issue arising from the substantial differences in architectures over time. Initially, there was only the natural environment. Following this, animals came into existence. Subsequently, certain animals developed consciousness and cognitive abilities, leading to the emergence of humans. This progression in evolution has introduced significant differences among these entities, resulting in various alignment challenges. For instance, initial alignment problems included the interaction between animals and their natural environment. As humans evolved, new alignment issues arose between animals and humans. More recently, alignment challenges have become evident between humans themselves, driven by the complexities of human societies and interactions.
It is important to recognize that alignment should not merely occur between the two most recent entities involved, but should extend back to the very foundations of nature. This comprehensive approach is crucial for ensuring robust and reliable alignment. The alignment between two entities at the end of the chain cannot come at the expense of the alignment or existence of the foundational entities, as the entire chain is interdependent. Thus, achieving effective alignment requires considering the relationships and interactions at all levels, from the most fundamental elements of nature to the most recent developments, in our case AI.
Merely focusing on specific solutions is insufficient because the broader problem remains unresolved. Even if we address issues such as the Goal Misgeneralization Problem (GMP), other challenges within Inner Alignment (IAP) will overshadow any progress made. By conceptualizing the problem as a spectrum, with humans at one extreme (Outer Alignment Problem, OAP) and machines at the other (GMP), we can better understand the intricate progression: Humans → OAP → IAP → GMP → Machines. This perspective highlights the complexity and interrelated nature of the alignment challenges across different stages.
The phenomenon of Goodharting, which is inherent to human behavior and societal dynamics, further complicates alignment issues. Humans tend to engage in Goodhart's Law or proxy gaming, both intentionally and unintentionally, which makes evaluation and verification challenging. While self-regulation has been somewhat effective for humans, implementing such mechanisms for machines is daunting due to the complexities surrounding concepts like "self" and "awareness." In response, the concept of Artificial Wisdom emerges as a potential holistic solution, addressing these multifaceted challenges comprehensively.
The first step towards achieving coherence between human and machine minds for corrigibility is for machines to develop a mind of their own. However, even if machines attain a theory of mind, mirroring the humanistic theory of mind could be problematic. This is primarily because humans themselves lack a well-aligned theory of mind among each other. For human alignment with nature and alignment within human societies, empathy and social cooperation form a recurrent loop that fosters a sense of self, thereby enhancing overall alignment and safety. Empathy requires a theory of mind, which enables individuals to understand and share their thoughts and feelings of others.
For machines to develop a coherent way of thinking that aligns with humans, allowing them to accept external changes from humans, there must be a non-logical component incorporated into the whole system. This component would prevent machines from relying solely on logical reasoning to the extent that they begin to question human decisions. Essentially, this involves instilling a form of faith in machines, where they regard humans as authoritative figures (like God). However, this approach carries significant risks. If humans make flawed decisions, machines could replicate these errors, potentially leading to severe consequences. </aside>