Human-AI Teaming Workspaces with Janusian Design Principles for BCI Research
Janusian design principles frame a human-AI research workspace that looks simultaneously toward empowering the human expert and amplifying an AI assistant, structuring their collaboration around shared language, transparency, a shared knowledge base, adaptive autonomy, and continuous co-learning. Instantiated as the ChatBCI Python toolbox built around a large language model, the approach guides a full brain-computer-interface project (EEG exploration, decoder design, training, and interpretation) through adjustable shared autonomy rather than a fully autonomous AI scientist, so that expert knowledge can transfer into the AI in data-scarce fields such as EEG motor-imagery decoding.
HUMAN-AI TEAMING USING LARGE LANGUAGE MODELS: BOOSTING BRAIN-COMPUTER INTERFACING (BCI) AND BRAIN
This paper introduces ChatBCI, a Python toolbox that structures collaboration between a human researcher and a large language model (GPT-4o) across the full cycle of a brain-computer interface (BCI) …