Towards Open-Ended Discovery for Low-Resource NLP

Natural Language Processing (NLP) for low-resource languages remains fundamentally constrained by the lack of textual corpora, standardized orthographies, and scalable annotation pipelines. While recent advances in large language models have improved cross-lingual transfer, they remain inaccessible to underrepresented communities due to their reliance on massive, pre-collected data and centralized infrastructure. In this position paper, we argue for a paradigm shift toward open-ended, interactive language discovery, where AI systems learn new languages dynamically through dialogue rather than static datasets. We contend that the future of language technology, particularly for low-resource and under-documented languages, must move beyond static data collection pipelines toward interactive, uncertainty-driven discovery, where learning emerges dynamically from human-machine collaboration instead of being limited to pre-existing datasets. We propose a framework grounded in joint human-machine uncertainty, combining epistemic uncertainty from the model with hesitation cues and confidence signals from human speakers to guide interaction, query selection, and memory retention. This paper is a call to action: we advocate a rethinking of how AI engages with human knowledge in under-documented languages, moving from extractive data collection toward participatory, co-adaptive learning processes that respect and empower communities while discovering and preserving the world’s linguistic diversity. This vision aligns with principles of human-centered AI, emphasizing interactive, cooperative model building between AI systems and speakers.

💡 Research Summary

**
The paper “Towards Open‑Ended Discovery for Low‑Resource NLP” argues that the prevailing data‑driven paradigm in natural language processing is fundamentally ill‑suited for the vast majority of the world’s languages, many of which lack written corpora, standardized orthographies, or any large‑scale annotation pipelines. While large language models (LLMs) have demonstrated impressive cross‑lingual transfer, their reliance on massive pre‑collected datasets and centralized compute makes them inaccessible to under‑represented communities, especially in the Global South. The authors therefore propose a radical shift: instead of training on static corpora, AI systems should acquire new languages through interactive, uncertainty‑driven dialogue with native speakers.

Core Contributions

1. Joint Human‑Machine Uncertainty Modeling – The paper introduces a composite uncertainty signal, U_total = α·U_human + (1‑α)·U_model, where U_model captures epistemic uncertainty (via Bayesian neural networks, deep ensembles, or entropy) and U_human is inferred from speaker hesitation, prosodic cues, conflicting corrections, and other meta‑linguistic signals. The weighting factor α allows the system to give more influence to either side depending on context.

2. Query Selection via Expected Information Gain – Given U_total, the system selects the next question Q* by maximizing expected information gain while penalizing human effort:
   Q* = arg max_Q E