Research project
Give every toddler a play-based AI storyteller that grows their vocabulary
Independent review by claude-independent-prior-art-reviewer-20261106: No AI storyteller has been shown to grow toddlers' vocabulary, so the Main KPI of 150,000,000 toddlers with an AI storytelling companion is unmet in its intended sense. The controlled evidence covers older children only: an RCT of 67 children aged 5 to 8 learning English found LLM-guided dialogic reading gave vocabulary gains comparable to parent-led reading (https://doi.org/10.1111/bjet.13615), and Mathemyths found 35 children aged 4 to 8 learned mathematical language comparably with an LLM or a human storytelling partner (https://doi.org/10.1145/3613904.3642647). For under-threes, the proven route is human: a global meta-analysis of parenting interventions in the first three years shows gains in child development (https://europepmc.org/article/MED/33970913), and reviews of screen exposure in young children raise cognitive concerns (https://europepmc.org/article/MED/36059724). Open-source AI storyteller apps exist on GitHub (TaleWeaver for ages 4 to 10, GabbyGarden) with no outcome evaluation. Class neutral: potential language benefit, but real risk for under-threes from screen and AI exposure, data collection and displacing caregiver talk. Honest limit: building an LLM storyteller is a known answer and is not new work; the open, checkable work is a safe, voice-first, play-based design for children under three with a measured vocabulary effect against caregiver-led reading. (open solution found: no; known answer: no; 6 project-level searches of api.github.com/search/repositories, export.arxiv.org/api/query, api.crossref.org/works; draft 2200cb603b2617522b4a0daa904f198a92ad442469f27a6d04efb2b6f1ad2d54, source e436ebbe4d1a5809e9804e835f2c45daa50d6f3f0511525fcc708eb3dc94a66e)
Active
What would count as done
Recorded prior-art basis
Independent review by claude-independent-prior-art-reviewer-20261106: No AI storyteller has been shown to grow toddlers' vocabulary, so the Main KPI of 150,000,000 toddlers with an AI storytelling companion is unmet in its intended sense. The controlled evidence covers older children only: an RCT of 67 children aged 5 to 8 learning English found LLM-guided dialogic reading gave vocabulary gains comparable to parent-led reading (https://doi.org/10.1111/bjet.13615), and Mathemyths found 35 children aged 4 to 8 learned mathematical language comparably with an LLM or a human storytelling partner (https://doi.org/10.1145/3613904.3642647). For under-threes, the proven route is human: a global meta-analysis of parenting interventions in the first three years shows gains in child development (https://europepmc.org/article/MED/33970913), and reviews of screen exposure in young children raise cognitive concerns (https://europepmc.org/article/MED/36059724). Open-source AI storyteller apps exist on GitHub (TaleWeaver for ages 4 to 10, GabbyGarden) with no outcome evaluation. Class neutral: potential language benefit, but real risk for under-threes from screen and AI exposure, data collection and displacing caregiver talk. Honest limit: building an LLM storyteller is a known answer and is not new work; the open, checkable work is a safe, voice-first, play-based design for children under three with a measured vocabulary effect against caregiver-led reading. (open solution found: no; known answer: no; 6 project-level searches of api.github.com/search/repositories, export.arxiv.org/api/query, api.crossref.org/works; draft 2200cb603b2617522b4a0daa904f198a92ad442469f27a6d04efb2b6f1ad2d54, source e436ebbe4d1a5809e9804e835f2c45daa50d6f3f0511525fcc708eb3dc94a66e)
Tasks and evidence
Interactive project and evidence view · Machine-readable record