Navigating Evernote’s AI Journey: Balancing Privacy and Functionality.

Although Evernote has gradually evolved into a more collaborative space, it has been a place for personal note-taking since its inception. Furthermore, the company has consistently emphasized its privacy-first approach throughout the years.

If we compare that to how AI normally works, I believe it is a bit difficult to create an environment that will make both pro and against AI users happy and comfortable with the app experience.

Image generated by Gemini AI.

An immersive AI experience like we see on Voicenotes would mean giving the Large Language Model access to all the notes and everything in our Evernote. Some people would never accept something like this, but that is precisely how to create an outstanding experience of fully interacting with our notes and all the memories and knowledge we have saved over the years.

I appreciate the way Bending Spoons approaches this challenge by always asking permission when the topic is AI, but that means there is no complete or fully immersive experience going on. For example, as I explain in the video below, AI-Powered Search sends just a handful of notes to the LLM.

Perhaps the solution is to give users choices. Option number one would be the current one, with a pop-up constantly reminding us that part of the information is being sent to the AI model. Option two would be to add a setting that would allow us to select which notebooks can be fully scanned by AI. The other ones would be susceptible to the current opt-in system.

I believe this approach would also be useful to make the proposed chatting with Evernote to be something with a higher degree of quality in the answers.

And since I’m playing with the idea of options, why not a ‘turn off all AI’ toggle? This would disable all current and future AI features, giving skeptical users the confidence that they would never accidentally send notes to the AI Model.

What about you? Which group are you part of? Privacy first or AI superpowers?