Paper proposes LLMs as algorithmic mediators in language evolution

A new paper proposes extending Salikoko S. Mufwene's ecological model of language evolution to account for the influence of Large Language Models (LLMs). The authors suggest that LLMs act as distributional mediators, altering the frequency with which linguistic variants are presented to human speakers through training and redistribution of outputs. This process, termed "algorithmic reweighting," can impact language evolution, though human social evaluation remains the decisive factor in whether model-associated forms become conventionalized or are avoided. AI

IMPACT This research could inform how we understand the long-term impact of LLMs on language development and human communication.

RANK_REASON The cluster contains a single academic paper discussing theoretical extensions to language evolution models. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

• alphaXiv
• arXiv
• CatalyzeX
• DagsHub
• Gotit.pub
• Hugging Face
• Large Language Models
• Salikoko S. Mufwene
• ScienceCast

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →