claim

active

claim:special-purpose-intelligences-optimized-for-narrow-tasks-may-not-converge-the-prh-only-holds-for-intelligences-performing-well-on-many-tasks

Special-purpose intelligences optimized for narrow tasks may not converge; the PRH only holds for intelligences performing well on many tasks

Key limitation of PRH

Source paper

extracted_from

The Platonic Representation Hypothesis

(2024) · Minyoung Huh · Brian Cheung · Tongzhou Wang · Phillip Isola

Neighborhood — ranked by edge-count

Quotes (1)

quote

Neural networks, trained with different objectives on different data and modalities, are converging to a shared statistical model of reality in their representation spaces.
contradicts
The paper's central thesis statement, presented prominently after the abstract

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.

Special-purpose intelligences optimized for narrow tasks might not converge to the platonic representationclaim0.890
Counterexample/limitation: only general-purpose models are subject to the convergence pressures described
Constraints and biases of human intelligence are best avoided by ‘designing for emergence’ rather than implementing a pre-existing theory of intelligence.claim0.762
Recommendation for creating non-anthropocentric machine intelligence.
Intelligence is the ability to achieve the same goal in multiple ways.concept0.761
If intelligence is defined by engaged concern for problem solving then the apparent limits of a system’s intelligence can be expanded by extending its sphere of concern.hypothesis0.756
Hypothesis linking care scope to intelligence ceiling.
What we really mean by 'human-level' intelligence is compatibility — a good impedance match between cognitive light cones — the ability to care about the same scale of goals and have the same existential struggles.claim0.754
Redefines 'human-level' AI from performance metrics to relational compatibility of cognitive scope
If intelligence is defined as engaged concern for problem solving, then the apparent limits of a system’s intelligence can be expanded by extending its sphere of concern.hypothesis0.753
Predicts that care-driven expansion of concern leads to higher intelligence.
Scale is sufficient but not necessarily efficient to reach high levels of intelligence; different methods can scale with different efficiency levelsclaim0.752
Implication of PRH for 'scale is all you need' argument
Curiosity, insight, decision-making, and diverse phenomena can all be accommodated by a single imperative: minimization of expected free energy (resolution of uncertainty).claim0.750
Central thesis of the paper unifying cognitive phenomena under one objective function