question

active

question:to-what-extent-can-physical-systems-learn-by-exploiting-typically-local-natural-processes-without-any-explicit-cost-function

To what extent can physical systems learn by exploiting typically local natural processes without any explicit cost function?

Central research question defining the scope of physical learning; asks about achievable learning under locality constraints

Source paper

extracted_from

Learning without neurons in physical systems

(2022) · Menachem Stern · Arvind Murugan

Neighborhood — ranked by edge-count

Claims (1)

claim

Physical systems are more constrained in learning abilities than in silico neural networks due to locality requirements, but this mirrors biological learning constraints and offers robustness benefits
gates
Core theoretical claim establishing that locality constraints in physical learning are not fatal—they reflect biological precedent and provide advantages like robustness and scalability

Related by similarity (8)

cosine ≥ 0.65 · no typed edge

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

Can Learning And Optimization Occur In Nonneural Physicalquestion0.772
All that is required is that this mechanical procedure is sensitive to the whole and is influenced by and guided by the structure of the whole. That is the secret of morphological emergence in natural phenomena.quote0.759
Load-bearing statement asserting that sensitivity to the whole is the essential requirement for generating novel living structure.
A living process is always natural, similar to the best kind of natural process, but it goes deeper and includes additional centers that naive natural process may miss.claim0.759
Positions living process as an refined version of innate human creativity, not an artificial imposition.
A physical learning system thus has the following ingredients: Physical d.o.f s that respond to external stimuli f by adopting a state or dynamic behavior s(f), a subset of which defines the desired behavior (i.e., output); Learning d.o.f wi that can modify how the physical d.o.f s(f; {wi}) respond to external stimuli f; A learning rule, dwi = h(s(f; {wi}))dt, that modifies the learning d.o.f wi based on the response of the physical d.o.f s to the stimuli f.quote0.758
Foundational definition of physical learning system components; load-bearing for understanding the entire framework
Different learning systems facing similar computational problems will converge to similar consciousness-like solutions, including potentially biological and artificial systemshypothesis0.758
Extension of the Universality Hypothesis to consciousness: if consciousness solves a well-defined computational problem, different systems will discover it independently
Evolution learns to generalize beyond default morphologies, producing problem-solving machines.claim0.750
Argues that evolutionary learning goes beyond specific adaptations.
The results generalise readily to non-equilibrium systems where scaling relationships remain similar (e.g., dynamic or localised scaling).claim0.749
Claim about broader applicability of the scaling argument
Learned simulations can be partially observed and lazily-rendered, and still work.claim0.746
One of the updates about prosaic ML simulation.