Barriers to Complexity-Theoretic Proofs that "AGI" Using Machine Learning is Impossible

Fuente: arXiv
Saved in:
Bibliographic Details
Main Author: Guerzhoy, Michael
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917444326522880
author Guerzhoy, Michael
author_facet Guerzhoy, Michael
contents A recent paper (van Rooij et al. 2024) claims to have proved that achieving human-like intelligence using learning from data is intractable in a complexity-theoretic sense. We point out that the proof relies on an unjustified assumption about the distribution of (input, output) tuples in the data. We briefly discuss that assumption in the context of two fundamental barriers to repairing the proof: the need to precisely define ``human-like," and the need to account for the fact that a particular machine learning system will have particular inductive biases that are key to the analysis. Another attempt to repair the proof, by focusing on subsets of the data, faces barriers in terms of defining the subsets.
format Preprint
id arxiv_https___arxiv_org_abs_2411_06498
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Barriers to Complexity-Theoretic Proofs that "AGI" Using Machine Learning is Impossible
Guerzhoy, Michael
Artificial Intelligence
Computational Complexity
A recent paper (van Rooij et al. 2024) claims to have proved that achieving human-like intelligence using learning from data is intractable in a complexity-theoretic sense. We point out that the proof relies on an unjustified assumption about the distribution of (input, output) tuples in the data. We briefly discuss that assumption in the context of two fundamental barriers to repairing the proof: the need to precisely define ``human-like," and the need to account for the fact that a particular machine learning system will have particular inductive biases that are key to the analysis. Another attempt to repair the proof, by focusing on subsets of the data, faces barriers in terms of defining the subsets.
title Barriers to Complexity-Theoretic Proofs that "AGI" Using Machine Learning is Impossible
topic Artificial Intelligence
Computational Complexity
url https://arxiv.org/abs/2411.06498