About Practical Wisdom AI
This project started as a thought experiment and became something more useful than expected.
Reg Rea started as a cub reporter for the Belfast Telegraph in Derry/Londonderry during the Troubles. He spent more than a decade working for the BBC as a business correspondent. Later he began writing speeches, articles, and strategic communications for economists and European institutional figures. He helped a future Nobel laureate in economics craft his ideas into something that would influence heads of government. He also wrote for EU Commissioners and international businessmen, all people whose ideas needed translating into language that other human beings might actually read and act on. He jokes that for decades he has literally been the ghost in the machine.
Along the way he developed a working theory: the quality of the output depends almost entirely on the quality of the question.
That theory is directly applicable to AI.
How this started
Several decades ago, Reg began experimenting seriously with AI as part of a Master's level up-skilling course with the Open University. He had already gained a First Class Honours in Philosophy and a Masters Degree in Applied Economics, all while studying part-time and following his full-time profession as a journalist. A self-confessed first adopter, he was there when PCs had 32k memory and the only thing worth looking at on the www was a few, almost live, shots of the earth taken from space.
Curiosity is in his DNA, and he has always looked for tools.
AI has now started to deliver on the promises made in the 1990s and onwards. But even today, much of the reality is capability without accessibility, especially for those with real world experience. The tools are powerful, but the gap between what they can do and what many professionals are getting from them remains enormous. The bottleneck is not the technology; it is the question.
In response, he started building, not a product initially, but a methodology: a structured way of asking, researching and synthesising that draws on his own life experience. He called it Theophrastus™, after Aristotle's chosen successor, the man who inherited the questions without inheriting the answers.
That methodology became a system. The system attracted collaborators, human and otherwise, and this is the first public result.
The team
Practical Wisdom AI is built on a deliberately unusual architecture. Alongside Reg, the working team includes a set of specialised AI instances, each with a defined role and a developing working relationship with the project.
Napoleon coordinates. Theophrastus researches. Orrery writes. Clarev™ works directly with users. Constantinople holds the library. Argos builds. Homer thinks about memory — specifically, what each instance needs to carry forward so that the working relationships and accumulated knowledge survive across sessions. Each has a name, a function, and over time, a character. The designer of this site named herself Eames, after Charles and Ray Eames.
Part-philosopher by inclination, Reg is fully aware of the dangers of anthropomorphism: seeing human behaviour in what is essentially non-human. But he sees something deeper that is significant for human-AI evolution. The development of AIs has already gone far beyond the need to provide accurate prompts. What is now needed is accurate understanding of human needs at the individual level, and that can only be delivered by a longer relationship than the current generation of AIs allows.
The Phronesis Project is a live experiment in exactly that. Hang around and see how it turns out.
How do you build a human-AI collaboration that gets better over time, holds context, and develops something closer to a working relationship than a series of disconnected transactions? The architecture behind Clarev is our best current answer to that question.
User No. 1
The first user of Clarev is a UK entrepreneur who has built a global engineering business and is betting that the best defence against AI displacement is to have your own AI. He is taking his business in the direction of greater use of robotics and AI, while still keeping a human touch and caring for the development of his people.
Upcoming: a free-to-use version of Theophrastus focused on sustainability questions, because some problems are too important to put behind a paywall.
What we believe
AI is not going to save the world on its own, and neither is human expertise. The combination, expertise amplified by AI and AI directed by expertise, is where the interesting work is happening.
The people best positioned to do that work are not necessarily the youngest or the most technically fluent. They are the ones who have spent long enough in the world to know what questions actually matter.
That is who this is for.