The Pope on A.I. – Let’s cut through the noise and build

The Pope on A.I. – Let’s cut through the noise and build

Pope Leo XIV has just published a 42000-word encyclical – a document intent to state and educate the Catholic Church of 1.4 billion believers – regarding the most disruptive technology so far I have experienced – A.I. Through this short write-up, I hope you can get some insights on how to build A.I. without sacrificing our future generations. The pope is inviting every readers to consider 1 question:

“What should we build?”

From a product builder point of view, this question is always oriented towards metrics such as user growth…etc. It may also translate to how powerful the A.I. can do the work of a human. None of the above qualities are not important. But it is missing something. Something transcends all technological advancement. Building for generations and for the common good of the society.

The Tales of Building Two Cities

The Pope permutated his whole encyclical with 2 biblical stories – the Tower of Babel and the rebuild of Jerusalem. When people are building the tower of Babel with pride and the pursuit of power, arguments and fights are ensured. God had to intervene out of love and separated our languages. In the other story, Nehemiah, the prophet that was tasked to rebuild Jerusalem took a different approach of synodality – discuss, listen and decide together. In the end, the people of Israel rejoiced that their city of God was rebuilt with their effort.

The contrast of the stories are so stark. But you may ask, so are those 3000 year old stories still relevant and give us insight today?

The resounding YES is echoing throughout the encyclical. The pope reflected on a lot of hidden costs in maintaining and developing A.I. and particularly LLMs. From the user experience to the data quality maintenance, and not to forget the environmental cost of running and building the hardware to run those models. I selected a few relevant points that I agreed upon. Also I will share my take on the implication on how we should build our AI product.

Our Standard in Building AI Agents

The first and foremost is “the disarming of AI”. AI is not for competing with humans and each other, but to free us to pursue our joy and happiness. It is also matched with how we are building our product to enable our users to do something they cannot do before and do their current job better, not to compete with us. All the projects we are working on is intended to build AI Agent that is of the following characteristics:

• Our clients have an option to self-host our solutions and remain sovereign to our clients and customers.
• Our solutions are never dictated and forced to our users. We co-create our solutions with all the stakeholders.
• AI Agents are doing all the repetitive and huge volume tasks, with a need of basic and non-critical classification
• AI Agents are not replacing the role and responsibilities of the current staff
• The current staff is always trained to monitor and steer the AI Agents.

The next important point Pope made is the culture of power and words have power to transform. AI agents are always good at data processing and mimicking human conversation. When we build our product, we build our AI Agents with both of them in mind.

• AI Agents only provide facts and options, never opinion and suggestion.
• AI Agents are writing and reading data in the same mechanism as a human staff.
• We stay in our lane of data storytelling and provide one of the all aspects of the reality

Something Beyond Technology

The next point Pope mentioned is more subtle but profound. It is the false ideologies of transhumanism or posthumanism. It basically means we are treating humans as an object that needs to be perfect and augmented. In our solutions, the AI Agent is fitting the people instead of the other way around. Concretely, we have some guidelines that might go against the current trends, we believe we should uphold nevertheless.

• Human is in charge of how the AI should behave and act
• We enable human rather than replacing them, or making them less skillful in their professional

The last point that is less relevant but still important to our solutions. The pope is aware of the “growing” vs the “building” analogy coined from the AI firms represented by Anthropics in the discussion. Some “behavior” of the A.I. is not fully understandable. The pope does not buy into it and urged the scientific community to dig deeper. We can also play our part by the following:

• We provide as much context and logs as we need to build towards explainability.

All in all, the document is aligned with our approach to work with our clients. We might be a small company but we will do our part to build a better future.

Thank you for reading so far and as a closing, I want to ask you one more time:

Do you want to build a Tower of Babel, or a new City of Jerusalem?

With Nous, it will always be a small part of the new Jerusalem.

— I am a father and a technologist who was born and raised in Hong Kong. I have worked in 3 continents to build systems and digital products with different sizes of companies, from seed stage startup all the way to MNCs. Currently I am working with Ben and Joseph to build AI Agents and other digital experience for non-profits and companies in the wellness space.