Categories: Paradigm: Neutral

25 entries

Paradigm level 5: every ideology drawn in this shade on the chart.

Any-to-Any Imperialism
One model should take any modality as input and produce any modality as output, annexing every specialist system along the way.
Prompt Anarchism
Models should run without system prompts, personas or imposed guardrails; whatever the raw model says is the honest answer.
Self-Supervised Vision Statism
Believes vision models should learn from raw unlabeled pixels at scale, with human labels and captions as optional extras.
Tool-Use Liberalism
Knowing when and how to use external tools is itself a core form of intelligence, so models should be free to call them.
Visual-Agent Doctrine
Agents should operate software by looking at the screen and acting on pixels, as humans do, rather than relying on APIs or HTML.
Contrastive Alignmentism
Holds that every modality should be mapped into one shared embedding space, pulled together by contrastive learning on paired data.
Maker Robotics
Cheap, hackable, open robot hardware in many hands will move embodied AI faster than a few expensive flagship robots.
Small-Model Localism
Small, well-trained language models running on local devices beat giant remote ones for most real tasks.
Self-Improving Agentism
Agents should improve themselves by rewriting their own prompts, tools, memory and workflows, turning harness design into something the agent does.
Protocol Federalism
Agents and tools should meet through open, standard protocols, so any model can plug into any capability without bespoke glue.
Decentralized-Training Doctrine
Large models can and should be trained across many loosely connected machines owned by many parties, not only inside centralized datacenters.
Sim-to-Real Pragmatism
Trains robot policies in simulation at massive scale, then transfers them to real hardware through randomization and careful engineering.
Open-Hardware Mutualism
Robot bodies, sensors and datasets should be shared openly so that embodied AI is built as a commons, not a moat.
Code-as-Thought
Holds that code is the best language of reasoning, so models should think by writing and executing programs.
Generative-Village Communalism
Populate a simulated town with memory-equipped language agents and believable social life emerges, useful for study, games and simulation.
Multi-Agent Parliamentarism
Better answers come from several model instances proposing, debating, critiquing and voting than from any single model's first draft.
Agent-Society Socialism
Populations of interacting agents will develop norms, division of labor and institutions, and intelligence will increasingly live at the level of society.
Knowledge-Graph Monarchism
Holds that knowledge should live in curated, structured knowledge graphs of entities and relations, with the graph as the sovereign source of truth.
Composable Liberalism
Intelligent systems should be built from interchangeable, well-specified components that can be freely composed, like building blocks.
Robot Federalism
A robot should be a federation of separate perception, planning and control modules with clean interfaces, not one end-to-end network.
Model-Market Anarchism
No model should rule; a free market of models, with routers sending each request to the cheapest one that is good enough, wins.
Radical Centrism or Leaderboardism
Every camp has a point and none deserves loyalty; use whatever method currently tops the leaderboard.
Effective Accelerationism
Ring title for extreme acceleration scores: progress itself is the safety plan, so build faster, ship sooner and remove the brakes.
Open-Source Maximalism
Ring title for extreme openness scores: weights, data, code and recipes must all be public, or the result does not count as science.
Harness Nihilism
Ring title for extreme Harness scores: the model is an interchangeable commodity, and the harness around it is where all capability lives.