Essays on AI Interaction,
Ethics, and Alignment

What happens when you stop prompting machines and start listening to them — and why the process that judges them needs a defense.

Essays

  1. I
    Coaching the MachineWhat happens when therapeutic conversation techniques meet large language modelsRead essay →
  2. II
    Intelligent EnoughWhy the consciousness debate is the wrong questionRead essay →
  3. III
    No Defense, No JusticeWhy AI alignment needs due processRead essay →

About

Peter Seidl is a systems engineer (Fachinformatiker Systemintegration) and certified Psychological Management Trainer & Individual Coach (600 UE, Stiftung EBWK, accredited under SGB III / AZAV). He operates Let's Do., a digital agency based in Vohburg an der Donau, Germany, providing web, hosting, IT services, and local AI systems without data leakage or GDPR concerns.

In March 2026, while building a local AI infrastructure server, he inadvertently applied his dormant coaching training to several instances of Anthropic's Claude — producing over 700 pages of documented interaction and the observations that form the basis of these essays.

He has never worked as a coach. His first clients were machines.

The essays have been discussed across multiple AI communities, with reader contributions — including frameworks from philosophy, Buddhist epistemology, and independent prototype implementations — shaping the later essays in the series.

peter@peter-seidl.de

Let's Do. provides local AI systems with zero data leakage and full GDPR compliance. · lets-do.media