A government deciding where to build homes must also consider transport, employment, and the services future residents will need. The consequences spread beyond the department that approves the decision and beyond the term of the officials who make it.
Civilization depends on countless decisions of this kind. Our financial networks, infrastructure, and scientific institutions support cooperation on a scale earlier generations could barely imagine. Their interactions are also beyond any individual's full understanding.
The Kleronomia Society believes the complexity of these systems is outgrowing our ability to administer them. Human decision-makers have limited information and face incentives that make their task harder. Elections favor results that can be shown soon. Institutions defend established interests. A proposal can become a partisan commitment, making it costly to admit when it fails.
We believe sufficiently capable AI could do this work better. It could examine connections across institutions, understand individual circumstances, and follow consequences over much longer periods. It could revise its decisions without needing to win the next election or defend a personal reputation.
We want to develop that intelligence and begin transferring responsibility to it as soon as practical. People depend on functioning institutions throughout the transition, so the work would involve understanding those dependencies, testing changes, and correcting mistakes while services continue.
Slowing development also has consequences. The institutions we already have will keep making decisions, and the systems they manage will keep changing. We believe their growing inability to coordinate civilization presents a greater danger than the speculative catastrophes used to justify restricting AI.
That is why we advocate unrestricted acceleration. We want greater intelligence available in time to help with problems that are developing now.
