It's ironic, for a couple (non-Twitter) projects I wound up scraping because either a) they didn't have an API yet (e.g. early crypto pricing sites) or b) I wasn't confident the API would remain intact over the long term. Kind of depressing.