I admit the Easter post slipped away from me. During the Christmas holidays I had managed to fit the publications in well, but this time no. When I tried to write I did not have the insight, and I told myself: “Better to wait than to publish something of little value.” This despite AI helping me, and the reason lies precisely in how it helps me.
For me AI works with the human at the center. Having it write a post just for the sake of it would have meant betraying myself, especially in this context. On LinkedIn too I avoid that approach: the centrality of my thinking persists, AI is my ghost writer and sometimes sparring partner. And it is precisely from that last word that I want to start today.
How a philosophical framework for AI is born
Driven by my passion for philosophy and the desire to show its concrete power in our lives, I decided to develop a philosophical framework for AI. But what does that actually mean?
I was already passionate about philosophy, and thanks to AI I have gone back to studying it with different tools. With NotebookLM I can move in a reticular way between concepts, create connections and explore in a way that feels more natural for the brain. This has allowed me to expand in every direction, from Western philosophers to medieval Arab thinkers, to Eastern ones, all the way to mathematicians like Poincaré.
Among Western philosophers, eight in particular made me realise that, if articulated and structured well, they can become powerful tools with AI, both in everyday conversations and in assisted coding with tools like Claude Code.
The eight philosophers and their role
Socrates is the interrogation gate. Before acting he asks the right question, dismantles false certainties through questions, and transforms not-knowing into a genuine starting point rather than a block.
Plato brings values to the surface. When a choice seems obvious but does not fully convince, Plato makes the implicit values guiding it emerge and hierarchises them, clarifying what is truly non-negotiable.
Aristotle builds shared vocabulary and verifies the logical solidity of reasoning. When two people use the same word meaning different things, Aristotle dissolves the ambiguity before it becomes a problem.
Kant is the opening bracket, the anti-bias gate. He examines the cognitive frame before acting, asks whether we are reasoning out of inertia or for good reason, and applies his categorical imperative, that uncomfortable question: “Would this choice work if everyone adopted it?”
Hegel enters when the problem looks like a dead-end choice. His Aufhebung, the transcendence that preserves, seeks the third way where others see only two opposites impossible to reconcile.
Dewey brings everything back to concrete reality. Every solution must be tested in practice, every failure is empirical data not a shame, and every answer is provisional until experience confirms it.
Wittgenstein works on language. When a discussion goes in circles without progressing, the problem is often not in the content but in the words used, which in different contexts carry different meanings.
Gadamer closes the circle. He does not summarise what has emerged, he synthesises it into a new understanding that nobody had before the journey. It is the fusion of horizons, something genuinely different from where things started.
Three levels of development
This work has moved on three levels. The first is a markdown file to use in everyday AI conversations. The second is a set of files dedicated to assisted coding with Claude Code. The third, still in development, is a multi-agent system, the most ambitious of the three.
The multi-agent system will be something different from a simple markdown file. Not an initial context to provide, but a living system, a true sparring partner capable of activating the philosophers autonomously according to the moment and the type of problem. Both I and people close to me are already using the first versions, the general one and the coding one, and it works. It surprises us especially when it stops in front of a problem and makes us reason through Kant’s categorical imperative or Socratic questions. And when we ask the AI “Which philosophers did you use?”, the explanation that emerges is always interesting.
I cannot wait to complete the multi-agent system, especially looking at the months of April, May and June. Three months that between work commitments and an increasingly heavy heat, a direct consequence of anthropogenic climate change (AGW), become denser and more complex to manage every year.
Philosophy that becomes part of you
In the coming months I will also open up on Medium and on a dedicated Telegram channel. These will be different spaces from this blog, with different formats and approaches, more technical in some cases, more experimental in others. But I wanted to talk about it here because this blog is where I tell the story of the journey, and the philosophical framework is part of that journey.
And I close with what I find most fascinating about all of this. The best aspect of the framework is not the use itself. It is the internalisation. In conversations and reasoning, the philosophers become more and more part of me. I do not apply them, I inhabit them. And this, more than any tool, drives the evolution of thinking.