
"The decision was prompted by Apple announcing its Journal app at WWDC that year. In that keynote, Apple said it would use "on-device machine learning" to provide prompts based on the content in your iPhone - things like contacts, photos, music, workouts, podcasts, and location data. The idea gave me the ick. Mainly because the app was described as a riff on the Memories feature in the Photos app, which at the time had "intelligently" resurfaced a photo of my mother's open casket."
"At my demo, Google told me the idea was to make journaling easier - much in the way that Gemini simplifies other writing tasks, like emails and document summaries. Sometimes, I was told, it can be hard to know what you should journal about. Looking back can also be difficult. The point of Gemini in this instance was to make life a little more convenient and helpful."
"I had flashbacks to that moment last week when I saw a demo of Google's take on its new Journal app. Except Google's Journal app leans harder into AI than Apple's version ever has. In addition to AI-powered journaling prompts, the on-device AI will also provide summaries of your entries. There's also a little calendar view that assigns a little emoji signifying your mood based on whatever you journaled that day."
Journaling apps increasingly rely on on-device machine learning and AI to generate prompts and summarize entries using personal data such as contacts, photos, music, workouts, podcasts, and location. Algorithmic features can surface sensitive memories unexpectedly and reshape the journaling experience with mood emojis and automated summaries. Convenience-oriented tools risk undermining the generative friction of confronting a blank page, which fosters deeper reflection, vulnerability, and unexpected discovery. Workflows that remove inconvenience can make journaling feel performative or curated rather than exploratory. Maintaining friction and manual effort preserves space for honest, slow thought and serendipitous insight.
Read at The Verge
Unable to calculate read time
Collection
[
|
...
]