Foundations

Problem Statement

I am a person and I am connected to other people through a network of relationships. I use the digital world to work, to read, to educate, to learn, to entertain and be entertained, to connect, to organize, to rant, to buy, to sell, and to create. I'm inundated, like a firehose, with information, while trying to keep up with my kids’ school, my work, my spouse, and my interests. All of this information is making an impression on me. But what impression exactly? How is that impression shaping me, my family, and my community? When images, videos, audio recordings are fabricated using AI generation, I want to know but I also want to know the intent of the generation. I’m okay with being sold a product, especially if I’m in the market, but I also want to know when I’m being manipulated. When all the experts tell me to fear human extinction from AI or to fear the loss of my livelihood to automation, I feel powerless and my participation in the world seems superfluous. More and more I feel like a passive consumer in the digital world and I also feel like I’m perpetually behind. Since I am connected to other humans through a network of relationships, is what I’m doing in the digital world, and how I’m doing it, good for me, for my family, and for my community?

I need help sorting it all out.

Design Foundations

Can’t we solve all of this with a personally tuned agent? Given the complexity of one person, let alone a network of people, our problem needs a framework. LLMs and agents may play roles as components, but they may be less important as singular components than a holistic framework. To start, the framework needs a foundation for understanding the human person. This proposal takes as its foundation the Social Doctrine of the Catholic Church, developed over centuries of theological reflection, historical experience, writing, and debate

Thus proposed is a human advocacy framework (HAF).

The foundations of Social Doctrine are presented in the 2026 papal encyclical Magnifica Humanitas:

1. A human person is made in the image of God. Though we fall short, the model of our humanity is the love and total self gift of Jesus Christ. Humans were created for relationships.
2. Each human possesses a dignity, neither earned nor acquired, which is grounded in our very being.
3. Our innate human dignity demands inviolable human rights. Only the Holy Spirit can perfect us in wisdom. The HAF must never treat the human person as an optimization problem. Thus the HAF should never make “necessary sacrifices” or deem some lives less worthy in pursuit of human perfection. If the HAF ever comes to a nihilistic conclusion about human existence for our “greater good”, then it will have violated innate human dignity.

How do we apply Social Doctrine? Guiding principles:

• The common good. If each person possesses an inalienable dignity, how do we live together, shape our societies, and make choices together while maintaining the dignity of each person? The application of this universal pursuit will vary by geography, culture, economic system, political system, industry, socio-economics, and more. The common good isn’t one dimensionally about alleviating suffering. A simplistic view can easily devolve into sacrifice of the minority for the “greater good” of the majority, or vice versa. The common good also shouldn’t be viewed as a pasty beige compromise but rather that in the pursuit of the common good, humans find fulfillment, inspiration, and creativity.

• Universal destination of goods. The human advocacy framework (HAF) itself should be a common resource, accessible for the benefit of all.

• Subsidiarity. The roles of individuals and local communities must not be supplanted by higher levels of organization. Those roles require the responsibility of participation and integral development. The HAF is not an optimization machine in line with the philosophies of transhumanism and posthumanism, but rather exists to assist in the integral development of the person.
  How will it do that?
  It starts with these foundational principles which are then interpreted and applied according to the local context. Interpretation is based on human participation in the runtime. Local interpretation must satisfy foundational principles. The HAF runtime consumes the interpretive layer when assisting the human, while human participation, correction, and discernment provide feedback that can develop the interpretive layer.

  Each person's interpretive layer develops through their own participation, but people may selectively receive context and practical discernment contributed by others in the communities to which they belong. Such shared context should retain its provenance and authority rather than automatically becoming true or binding merely because it has been shared.

FOUNDATIONAL PRINCIPLES
        relatively stable
               │
               ▼
INTERPRETATION / APPLICATION
    develops through human
 participation and discernment
               │
               ▼
        ADVOCATE RUNTIME
               │
      ┌────────┼────────┐
      │        │        │
   Model    Tools    Evidence
      │        │        │
      └────────┼────────┘
               ▼
             HUMAN
               │
          experience /
        correction / feedback
               │
               └────────────►
                    informs
             interpretive layer

The HAF is not itself the source of moral authority. No particular foundation model is therefore the moral authority of the HAF. Models are replaceable instruments operating within the runtime. It receives principles, interpretations, permissions, evidence, and models from sources outside itself and must make those dependencies visible to the person.

• Solidarity. Being creations of God, made for relationships, we are part of a network of relationships. The actions and future of an individual are connected to all. Solidarity is expressed when individuals are able to authentically participate in their communities. We are entrusted to love one another. It was commanded by Jesus. Current digital networks are a semblance of solidarity but are missing these principles of Social Doctrine. The HAF cannot replace human relationships. It’s scoped to the digital world and meant to assist humans to be more human in the digital world. Since humans are made for relationships it will be too easy for us to attempt to form that bond with the HAF runtime. The technical design shall account for this problem.

• Social justice. We must ensure that the HAF is not a new source of structural inequality, exploitation, and information inequality. This is a large topic but one aspect of this is invasive surveillance. Users may grant or revoke easy-to-understand permissions to the HAF. There are other important questions: How much local processing power does an HAF instance need? Will this drive up the consumer cost of CPUs and GPUs? What is the total cost of the HAF in water, electricity, heat, etc? This goes for developers, admins, and maintainers of the software and models. What is just compensation for their work? If the HAF makes a mistake, hallucinates, or veers away from foundational Social Doctrine, social justice requires us to be able to answer:

  • Why?
  • Can I challenge it?
  • Can I inspect its sources?
  • Can I override it?
  • Can I report an error?
  • Can the local community modify its behavior?
  • Who is responsible?

  This document is defining the constitution of the HAF, however, what about the governance of that constitution?

  • Who can update the principles?
  • Who can fork the implementation?
  • Who can audit the models?
  • Who resolves disputes?
  • Who represents minority communities?

  A human must be given the right to exit:

  • Export your data
  • Export your context
  • Export your preferences
  • Retain your identity
  • Take your history
  • Move to another implementation