A baby hears thousands and thousands language models before they say their first word! We need to look at typical language development as it applies to our AAC users. We cannot just place a speech-generating device in the hands of a student with the expectation that they will know how to use it by osmosis. Research shows that modeling language on a device, without the expectation/demand for a reply, is a great way to increase skills. Modeling takes time. We need to be patient. Our student may not independently use their device until they have seen/heard thousands of words modeled for them.