Type I, Planetary
A planet becomes
a system.
Order of magnitude
1016 W, planetary energy
- energy
- planet
- coordination
- global
- automation
- networked
- infrastructure
- orbital
Planetary grids, climate systems, orbital industry.
Type II, Stellar
A star becomes
infrastructure.
Order of magnitude
1026 W, stellar energy
- energy
- star
- coordination
- systemwide
- automation
- self-directed
- infrastructure
- swarm
Dyson-swarm stellar industry.
Type III, Galactic
A galaxy becomes
reachable.
Order of magnitude
1036 W, galactic energy
- energy
- galaxy
- coordination
- distributed
- automation
- self-replicating
- infrastructure
- interstellar
Self-replicating galactic infrastructure.
Research notes
Questions at
the threshold.
Definitions, project context, and the proposed research pathway.
Kardashev on XWhat is effective acceleration?
Effective acceleration means shortening the time, coordination, and engineering effort required to cross civilization-scale energy thresholds. AI does not create extra energy. It may help people model complex systems, coordinate work, discover better designs, and deploy infrastructure more efficiently.
What is the Kardashev scale?
The Kardashev scale classifies civilizations by the energy domains they can use. Type I operates at roughly planetary scale, around 1016 watts. Type II reaches stellar scale, around 1026 watts. Type III reaches galactic scale, around 1036 watts.
What is Kardashev Research AI?
The project describes Kardashev Research AI as a nonprofit research initiative by @beffjezos. Its stated purpose is to explore how AI can accelerate civilization-scale science, engineering, coordination, and infrastructure.
What are $E/ACC and $KAI?
$E/ACC is the community token intended to contribute to public funding that makes Kardashev AI research possible.
CbcyNo7m1amFWqEQm2m4PLv1UNvpcL3C1Ujm6AkzpKoU
$KAI is the planned economy token for research ecosystem activities. Its intended or prospective utilities may include compute credits, research grants, contributor rewards, open bounties, dataset and model access, licensing settlement, experimental-market participation, and community governance. These utilities are plans, not guarantees.
soonRoadmap
The proposed pathway begins with foundational models, simulation, and research tooling. It then advances toward specialist AI systems for science and engineering, followed by consumer products that translate research into usable tools. The longer-term aim is interoperable agent and model infrastructure intended to accelerate AGI research, without assuming that AGI is guaranteed or imminent.