There are many billions more waiting to be curated and digitized in collections around the world. See also GBIF, for tons of data on the tiny fraction that has been digitized. There are > 2 billion records there alone.
And now for a shameless plug. If you want to help one of the handful open-source tools (there are numerous commercial tools used as well) that are used to digitize these materials so that you can access all that goodness please check out TaxonWorks (https://taxonworks.org). We don't serve many collections, but we're growing, and doing some cool things. As a hint of scope of the problem, for those handful of collections we do serve, we have ~1 million specimens digitized from an estimated total of well over 20 million individuals. Lots of cool things to consider throughout, including this "AI" stuff.
And now for a shameless plug. If you want to help one of the handful open-source tools (there are numerous commercial tools used as well) that are used to digitize these materials so that you can access all that goodness please check out TaxonWorks (https://taxonworks.org). We don't serve many collections, but we're growing, and doing some cool things. As a hint of scope of the problem, for those handful of collections we do serve, we have ~1 million specimens digitized from an estimated total of well over 20 million individuals. Lots of cool things to consider throughout, including this "AI" stuff.