I believe everything that can be done in a computer can be automated

My aim is to automate knowledge work. I believe that everything that can be done in a computer can be automated, and I want to find out how far that goes. Not as a claim about today, but as the direction I keep building in.

The next step, as I see it. I think the earlier steps that took work away from people also gave people somewhere else to stand, and I hope this one does too, if the person stays the centre of it: the machine does the work, the person decides what is worth doing and says yes. That is the shape I try to build, and this page is me finding out whether it holds.

The risks, named. Dependence: when the factory does the engineering, I lose the hands that did it. Attention: a system that handles everything can also take everything, so I keep every decision that matters as a card I must answer. Who benefits: automation inside a vendor’s product benefits the vendor first. I build this for one person, so at least in this case the answer is clear.

Not AI inside silos. I do not believe in AI that lives inside one product. It serves that product’s vendor, not the person who has to jump between ten panes. A silo does not meet the need of me as a person.

One pane for one person. I need one place where I can reach and use everything, and the AI should sit there, with me, not inside each tool. The tools can stay where they are. The pane is mine.

The products behind are stores of logic and data. I see the systems behind as repositories: the business logic for one process, and the database for that process and its data. My process spans many systems, and therefore many databases. The process is mine; the systems are where it keeps its things.

A user of one. Nobody but me is focused on me. In every product I am one more user. Pharos is built around one person, and that is me, as the user and not as a technician.

Installable anywhere, as a by-product. The platform is built to sustain itself and to be installed into any organisation. That is not the goal. It is a side effect of what interests me: how to create systems that work as a whole.

Think big, deploy fast. I cannot stop thinking bigger; it is part of who I am. It does not stop me from deploying fast. The two go together: the big idea sets the direction, and the small, frequent deploys are how I find out whether it holds.