The Masters Thesis
This is the working map: the claims we keep returning to, the conversations that sharpen them, and the questions that still refuse to settle.
Intelligence will distribute, not just centralize.
The future of AI will not be one system running the world. It will be many intelligences: personal, institutional, local, frontier, open, closed, and negotiated through protocols.
Bigger centralized models keep winning in capability. Ownership, privacy, latency, cost, and resilience keep pushing intelligence outward.
Intelligence is becoming abundant. The scarce things are context, trust, taste, accountability, and agency.
Reading path
The thesis is built through essays, not slogans. Start with the pieces that carry the most weight.
- 01
The Abundance Question
The philosophical stake: who we become when work and scarcity change.
- 02
The Internet of AI Agents
The infrastructure stake: how agents discover, verify, and act.
- 03
Ontology, Context and Semantic Data Layers
The enterprise stake: how companies make themselves legible to agents.
- 04
What the Model Thinks But Doesn't Say
The trust stake: what changes when model internals become inspectable.
Still unsettled
A good thesis should sharpen the questions it has not answered yet.
- 01
Can personal agents become a real ownership layer, or will they collapse into platform-controlled assistants?
- 02
Do skills replace apps, or become the new packaging layer inside existing software businesses?
- 03
What does accountability mean when an agent acts across company, customer, and vendor systems?
- 04
Which forms of work become more valuable when intelligence is cheap?
- 05
Who gets to inspect, govern, and shape the systems that increasingly act on our behalf?