[codeulike]

Does it have language skills? I always thought it would be interesting to train a model within minecraft as a sortof proxy for 'embodiment'. You could then try asking it about its experiences. "whats your favourite food?", "How does it feel when you hear a spider?", "how low does your food bar need to go before it feels really urgent?" etc

[agajews]

Not yet, but we'll add language as a modality to the larger models! The models are trained end-to-end on video data, so we'll need datasets that mix video and language, e.g. transcripts of game streams. When the models are scaled up to cross-videogames and robots there will definitely be a bunch of language data.