During Google I/O 2023, Google announced that it would be injecting AI into its Messages app by way of a feature called “Magic Compose”. This is a part of the company’s larger AI integration initiative, which has already begun transforming various aspects of its services. From the “Help me write” feature in Google Docs to the experimental chatbot Bard (which I’m still very unimpressed with), Google continues to push the boundaries of what AI can do, all while claiming to be responsible in the process. Whether or not that’s true is still yet to be seen.
To try this new feature out, you’ll need to become a part of the Google Messages beta program over on the app’s Play Store listing. Once enrolled, you can enable the Magic Compose feature specifically for RCS chats within the app’s settings. That’s right, it only works for RCS at this time. There’s something else you should know though – it’s also being prioritized for Google One subscribers with a 2TB+ plan ($9.99 USD per month) or higher since it’s a part of the larger “Search Labs” opt-in feature which is also barred behind that tier.
However, as more “spots become available”, this will roll out to other users who don’t meet the three qualifications – an appropriate Google One plan, a Google Messages beta enrolled user with RCS Chats enabled and an appropriate Google One plan. Oh, and as you may have suspected, this is U.S. only, in English, and not for anyone rocking non US SIM cards.
With all of that out of the way, Magic Compose can generate or alter your message drafts in one of seven styles – Remix, Excited, Chill, Shakespeare, Lyrical, Formal, or Short. All you have to do is tap the magic pencil icon at the right of the text entry box and choose one of the aforementioned styles. I think would be kind of funny to send someone messages as Shakespeare and to keep doing so for a month straight to get their reaction, but the person on the receiving end may not agree.
According to Google, it won’t store or save these generated messages, especially as your chats are end-to-end encrypted, so they’re created on the fly and only exist on your device and are only sent to the person you’re conversing with. Let’s discuss – would you use Magic Compose or would you feel inauthentic?
How long before we aren’t even sure if the person writing the email or sending a message is actually doing those things using their own mind? Google says that AI should be used as a “springboard” for your ideas, but how long until it gets good enough to well, not need us to do the thinking after the draft is created?