A Python implementation of David Deutsch’s Constructor Theory framework, exposing key concepts—from simple Tasks and branching substrates to quantum-gravity and electromagnetism—entirely in code. Includes a “universal constructor” that can bootstrap itself from a list of Tasks, demonstrating self-replication and the power of Constructor Theory.
“A demonstration of how constructor theory could be explored in code, not a high-precision physics engine. For the formal definitions, see David Deutsch and Chiara Marletto’s recent paper “Constructor Theory of Time” (May 13, 2025).
-
Core framework: Attributes, Substrates, Tasks, Constructors
-
Irreversible & quantum tasks: Many-worlds branching, decoherence guards
-
Timers & Clocks: Simulate proper-time, special/general relativity corrections
-
Fungibility & SwapConstructor: Free exchange of identical substrates
-
ASCII visualizer:
ascii_branch()for quick text-based branch inspection -
Continuous dynamics: 1D & 2D substrates,
DynamicsTask, RK4 & symplectic integrators -
Coupling tasks:
- Gravitational two-body (1D)
- Coulomb coupling (1D)
- Lorentz-force (2D) for charged particles in a magnetic field
-
Quantum-Gravity & Electromagnetism: Graviton & Photon emission/absorption Tasks
-
UniversalConstructor: Bootstraps any list of Tasks into a working Constructor
-
Pluggable Backend Architecture: Modular task ontologies for different physics domains
-
Hydrogen atom constructors: Excitation, deexcitation and two-atom collisions
-
Demo scripts:
demo.py– shows every constructor in actionbootstrap_demo.py– elegant self-replication via the UniversalConstructorbackend_demo.py– demonstrates the pluggable backend system
- Python 3.8+
- (Optional)
matplotlibfor phase-space plots
git clone https://github.com/gvelesandro/constructor-theory-simulator.git
cd constructor-theory-simulatorpython -m unittest ct_tests.pypython demo.py
python bootstrap_demo.py
python backend_demo.pyNote: If you don’t have
matplotlib, the demos will still run; plots will simply be skipped with a warning.
from ct_framework import (
Attribute, Substrate,
PhotonEmissionTask, PhotonAbsorptionTask,
UniversalConstructor, ascii_branch
)
# 1) Define your “program” of photon Tasks
ELEC = Attribute("charge_site")
prog = [
PhotonEmissionTask(ELEC, emission_energy=5.0, carry_residual=False),
PhotonAbsorptionTask(ELEC, absorption_energy=5.0)
]
# 2) Build a Constructor at runtime
uc = UniversalConstructor()
em_cons = uc.build(prog)
# 3) Emit a photon
atom = Substrate("A", ELEC, energy=20.0)
branches = em_cons.perform(atom)
print(ascii_branch(branches))
# => * charge_site (A)
# * photon (A)
# 4) Absorb it back
photon = next(w for w in branches if w.attr.label=="photon")
restored = em_cons.perform(photon)[0]
print(restored)
# => A:charge_site(E=20.0,Q=0,t=2,F=charge_site)The framework now supports a modular backend architecture for different task ontologies:
from ct_framework import (
UniversalConstructor, Substrate, Attribute,
ElectromagnetismBackend, QuantumGravityBackend
)
# 1) Use individual physics backends
uc = UniversalConstructor()
# Pure electromagnetism
em_constructor = uc.build_from_backends(["electromagnetism"])
# Pure quantum gravity
qg_constructor = uc.build_from_backends(["quantum_gravity"])
# 2) Combine multiple backends
unified_constructor = uc.build_from_backends([
"electromagnetism",
"quantum_gravity",
"hydrogen_atoms"
])
# 3) Create custom backends
class CustomPhysicsBackend:
def get_tasks(self):
return [/* your custom tasks */]
def get_name(self):
return "custom_physics"
# 4) Available built-in backends:
# - "electromagnetism": photon emission/absorption
# - "quantum_gravity": graviton emission/absorption
# - "hydrogen_atoms": H excitation/deexcitation
# - "continuous_dynamics": integrator tasksRun python backend_demo.py to see the full backend system in action!
This is intended as an educational resource and proof-of-concept. Contributions are very welcome! Please:
- File issues for missing tasks or physics modules
- Submit pull requests for new constructors (e.g. chemical reactions, friction)
- Improve documentation or add more demos
Released under the MIT License.
- Inspired by David Deutsch and Chiara Marletto’s work in Constructor Theory and their May 13, 2025 paper “Constructor Theory of Time.”
- Thanks to the quantum-foundations community for feedback and discussion.