Are there Mac OS so-called Smart Compose (Real-Time Assisted Writing) apps similar to what gmail offers?
I am referring to suggestions to finish sentences (which can be implemented by pressing the tab key) , not auto complete which corrects words.
For those that want to help, but don’t know the Smart Compose feature, Google support sez:
“You can let Gmail help you write emails faster. The Smart Compose feature is powered by machine learning and will offer suggestions as you type. …. Smart Compose personalized suggestions are tailored to the way you normally write, to maintain your writing style. Only you see your own private, personalized suggestions for your account. No other users, including administrators for your organization, can see your personalized suggestions. When personalization is turned off, you see generic suggestions as you type. …. Smart Compose is not designed to provide answers and may not always predict factually correct information.”
It was apparently trained on lots of corporate email communication data.
Unless you are using Scrivener for writing business memos, it is hard to see how a “sentence finisher” system like Smart Compose could be useful in the context of Scrivener. And I seriously doubt that training such a system up on a body of fictional literature would do anything sensible for your writing process — assuming you have your own ideas about where your story is going!
This is not to say it wouldn’t be fascinating to toy with a system so trained for fiction. I suspect we would all want to try it, and none of us would use it when actually writing fiction!
I write neither fiction nor business memos.
I have experience with and published on the topic of neural networks. I have seen first hand their ability to learn based on experience and make predictions In this case based on the individual’s writing history and pattern, irrespective of what they are writing about.
Jasper is an example, but the cost is prohibitive
see my reply below. If that’s the way you write, that’s what the neural network will base its predictions on. You could write in Papua New Guinean, if you write enough to train the neural network. That’s what deepl, the best translation software available IMO is based.
I know. Was just curious about the implications. The assistant predicts what you would write in a comparable situation and how. So you don’t have to do it. It will sound like you, in the past. It can’t come up with original ideas. But if it could, based on your way of thinking and way of talking: Who did actually write it?
On a more serious note: I’m not aware of a free or reasonable priced option with tight macOS integration. Would totally be interested to toy around with it, though.