Gemini’s scheduled activities changed the way I use apps

Although I’ve used the Gemini app for a variety of tasks from restoring images to suggesting new music, I didn’t realize how useful scheduled actions could be until I tried them. As someone who is incredibly forgetful, getting an AI to perform certain tasks without manual input has made a range of tasks easier for me.

Scheduled actions in Gemini are automated prompts, which allow the chatbot to generate responses at specific intervals. For example, you can get it to generate weekly or monthly reports, summarize daily news, or provide updates on your schedule.

However, access to this feature depends on your account. If you have a personal account, you can only use the functions set on the Google AI Pro or Ultra plans. For work and school accounts, the feature is limited to specific Google Workspace editions. You will also need to sign in to Gemini Apps.

If there’s a specific prompt you’d like to repeat at regular intervals, you can set it in a chat with the AI. I usually start with the sign I want so I can see how Gemini will react, making changes to the instructions if necessary. Then I ask her to repeat this action – for example, tell her to “repeat this every Tuesday at 10 am.”

If you use the Gemini mobile app, you will receive a push notification when your scheduled action is ready. Otherwise, the newly created response will appear at the top of your chat. You can see that new content has been added based on the blue dot that appears next to the chat.

When I first set up a scheduled action, it was more out of curiosity than anything else. However, after weeks of using this feature, it has saved me a lot of time by automating some tasks that I usually forget. This has become one of my favorite Gemini tricks.

The first task I set up was to produce a weekly report of consumer technology news, citing sources. Although keeping up with technology is part of my job, I stay away from announcements outside of my work hours because I’m usually completing work or trying to relax.

This means I sometimes miss new features or products that I’m not directly covering as part of my assignment. Relying on Google Discover isn’t working well either, because it keeps me in a bubble.

To fix this, I started getting Gemini to compile a summary of last week’s announcements. This has helped me stay on top of news that I may have missed. Over time, I’ve changed the report to include topics like open-source software, changes in interoperability for major platforms, and any new trends.

Although I still have to spend a lot of time reading news and keeping up with software changes, Gemini’s reports help address any unknown points I have. Since I focus most of my work on Android, it has kept me informed about changes and announcements to Apple products.

The second automated routine I love is my weekly series recommendations. I’ve been in too much pain to pursue my gaming hobbies for the past few months, so I spend most evenings on the couch watching series. As a result, I run out of things to watch very quickly, especially since I can be a bit picky.

Tired of reading dozens of articles on what to stream, I decided to set up a scheduled action for weekly recommendations. The advantage is that I can tell the AI ​​what genres I like, provide a list of shows I’ve already watched and liked, and ask for new recommendations based on those preferences.

Whenever Gemini suggests something that sounds interesting, I can ask to add that show to a watchlist that it saves in Google Keep. It automatically updates this list when I ask it to add a new item, making for an extremely intuitive process.

Part of me wishes this feature was available on the free versions of Gemini with limits on the frequency of scheduled tasks. After all, free users have access to a limited number of signals using Gemini’s Pro model.

The big upgrade that Scheduled Actions provides is the convenience feature, which is useful if you’re forgetful like me.

It’s possible to set up automation using a platform like Zapier, but it limits you to specific models, which impacts output. However, if you want to tinker with automation without the Google AI Pro plan, this is one way to do it.