Note: This paper was written as a thought piece in 2023 when the price of small robots was too high for government-sponsored research along these lines. As the cost of child-size robots comes down, substantial research projects in this field may become more feasible along the trajectory suggested in this paper.

Developmental AI is a research field that seeks to create embodied AIs that develop human-like abilities. These AIs would start with innate competences and learn more by interacting with the world including people. To date, some developmental AIs have been demonstrated, but their abilities so far do not surpass those of pre-toddler children.

In contrast, mainstream approaches have led to impressive feats and commercially valuable AI systems. The approaches include deep learning and generative AI (e.g., large language models) and manually constructed neurosymbolic modeling. However, manually constructed AIs tend to be brittle even in circumscribed domains. Large language models (LLMs) are helpful on average, but they can make strange mistakes and not notice them. Not learning from their experience in the world, they can lack common sense and social alignment.

This position paper lays out prospects, gaps, and challenges for a bootstrapping approach to developmental AI that follows a bio-inspired trajectory. The approach creates experiential foundation models for human-compatible AIs. A virtuous multidisciplinary research cycle has led to developmental AIs with capabilities for multimodal perception, object recognition, and manipulation. Computational models for hierarchical planning, abstraction discovery, curiosity, and language acquisition exist but need to be adapted to an embodied learning approach. The remaining gaps include nonverbal communication, speech, reading, and writing. These competences enable people to acquire socially developed competences.

Aspirationally, developmental AIs would learn, share what they learn, and collaborate to achieve high standards. They would learn to communicate, establish common ground, read critically, consider the provenance of information, test hypotheses, and collaborate. Success in the developmental approach would make the training of AIs more democratic.

Publications

Stefik, M., Price, R. (2023) Bootstrapping Developmental AIs (112 pages) arXiv  http://arxiv.org/abs/2308.04586

(The featured image of a baby robot stacking blocks was generated by Perplexity.)