The backdrop of the shenanigans of ChatGPT, Bard and other artificial intelligence (AI) systems these days has only served to accentuate how increasingly frustrating working with Google Docs is. I use Docs every day to write my articles and edit those that the freelancers I’m working with have filed. I don’t use tools like Grammarly but I do pay attention Docs’s blue and red underlines indicating grammatical and typographical aberrations, respectively. And what Docs chooses to underline either way is terribly inconsistent. I have written previously on how Docs ‘learns’ grammar, based on each user’s style, and expressed concern that its learning agent could be led astray by a large number of people, such as Indians, using English differently from the rest of the world and thus biasing it. Fortunately this issue doesn’t seem to have come to pass – but the agent has continued to be completely non-smart in a more fundamental way. This morning, I was editing an article about homeopathy on Docs and found that it couldn’t understand that “homeopathy”, “homeopathic”, and “Homeopathy” are just different forms of the same root word. As a result, correcting “homoeopathy” to “homeopathy” didn’t suffice; you have to correct each form to remove the additional ‘o’.
It gets worse: the same word in bold is, according to Google Docs, a different word…
… as is the word with a small ‘H’.
Google has a reputation for having its fingers in too many pies and as a result neglecting improvements in one pie because it’s too busy focusing on another. There is also a large graveyard of Google products that have been killed off as a result. There’s some reason, for now, to believe Docs won’t meet the same fate but then again I don’t know how to explain the persistence of such an easily fixable problem.