Dictionary
Concept dictionary
The vocabulary you need to read this archive. Each entry is the term as the field uses it, paired with a paper in the corpus where Wissner-Gross works with it.
The definitions are ours, written to get you into the papers — they are not quotations of Wissner-Gross. The citation under each one is the part that is his.
- Causal entropic force
- A force, in the thermodynamic sense, that pushes a system toward states from which the greatest diversity of futures remains reachable within a finite time horizon. Proposed here as a formal physical relationship where earlier work had only "hinted at a possible deep connection between intelligence and entropy maximization".
- In the corpus Causal Entropic Forces (2013) ↗
- Causal path entropy
- The entropy computed not over the states a system might occupy now, but over the paths it might still take between now and a horizon. Maximizing it is what the causal entropic force does; the horizon is the parameter that makes it finite and computable.
- In the corpus Causal Entropic Forces (2013) ↗
- Time horizon
- How far ahead the futures are counted. It is the free parameter of the whole construction: too short and the behaviour is myopic, too long and it is intractable — the reason this archive’s AI host is called Horizon.
- In the corpus A Thermodynamic Model for Behavioral Intelligence (2012) ↗
- Dielectrophoresis
- The motion of a neutral particle in a non-uniform electric field, caused by the field inducing a dipole in it. Used in this corpus to assemble, reconfigure and disassemble silicon nanowires between electrodes — building structure with a field instead of a tool.
- In the corpus Dielectrophoretic reconfiguration of nanowire interconnects (2006) ↗
- Self-assembly
- Structure that forms from local interactions between components rather than from external placement. The bench work and the cellular-automata work are two views of the same interest.
- In the corpus Dielectrophoretic control of semiconductor nanostructure growth and assembly (2005) ↗
- Cellular automaton
- A grid of cells updating in lockstep by a local rule — a minimal model of pattern formation. This corpus takes the unusual step of studying *ensembles* of rules rather than single ones, because real biological and physical systems coexpress many rules at once.
- In the corpus Pattern Formation without Favored Local Interactions (2008) ↗
- Rule ensemble
- A collection of cellular-automaton rules applied together rather than one rule applied everywhere. The companion paper reports what it calls "the first known analysis" of the equally-weighted ensemble, where no local interaction is favoured — and finds pattern anyway.
- In the corpus Pattern Formation without Favored Local Interactions (2008) ↗
- Relativistic statistical arbitrage
- Trading two securities on geographically separated exchanges when the light-propagation delay between them is no longer negligible — so the optimal place to run the strategy is a physical location determined by the speed of light.
- In the corpus Relativistic statistical arbitrage (2010) ↗
- Participatory telerobotics
- Treating people carrying mobile devices as ad-hoc telerobots — generalising participatory sensing to include real-time teleoperation and telepresence.
- In the corpus Participatory telerobotics (2013) ↗
- Few-shot learning
- Learning a new task from very few examples. Invoked in this corpus for autonomy in contested environments, where prior approaches were limited by "the low survivability of available AI and machine-learning techniques in the face of novel situations".
- In the corpus Adaptive Online Learning for Human-Robot Teaming in Dynamic Environments (2019) ↗
- Variational graph autoencoder
- A generative model that learns compressed representations of graphs — entities as vertices, relationships as edges. The 2021 paper adds a sparsity bias so the learned representations suit sparse graphs.
- In the corpus Joint sparsity-biased variational graph autoencoders (2021) ↗
- Physical Superintelligence
- The company he founded and serves as Chief Scientist of, per his own site. The name is worth reading against this corpus: two decades of asking what physics already implies about intelligent behaviour.
- In the corpus alexwg.org ↗