Power to the users: feature toggles

A feature toggle is a non-visible option to enable or disable. In software development, they are commonly used to enable features without the need of an update for a group of users, testers or internal developers to be able to use (new) functionality before it is fully deployed to everyone. This is useful in many scenarios:

• enable a new feature only for beta testers with the possibility to add more and more non-beta testers without another rollout.
• let developers and support teams use debug information to understand bugs
• release half-finished features to be able to test in the real world without any user noticing
• rollback a feature after negative user feedback or when something does not work as expected

There are even SaaS companies that specialize on feature toggles.

Today, I want to write about even bigger feature toggles. Feature toggles to (mostly) disable bigger features. It is almost impossible to disable features such as Siri, Cortana, or Bixby, any AI feature (Apple Intelligence, Samsung Image AI, Google Gemini). All these features are deeply integrated in the mobile or desktop operating systems.

Once in a while, I get ask to enable Siri, e.g. by accidentally clicking the Siri button on the MacBook TouchBar, by pressing the Siri shortcut or by selecting the wrong button after a macOS major update. I have Siri disabled on my Mac, yet it is present all the time and ready to be used. I simply want to get rid of it altogether.

With the hype of AI in the last few years, you now can find it everywhere: each smartphone has an integrated AI image editing tool, AI image search, AI assistants such as Gemini, AI writing assistance and autocompletion, “smart” notifications, and many more. Even operating systems have it integrated: MS Windows Recall, macOS Apple Intelligence, and browsers integrate it aggressively, with Google adding Gemini more and more into Chrome. There are also all these apps that have some “intelligence” features, such as sorted emails, photo editors, intelligent search or recommendations (booking.com, Airbnb, etc.) or chatbots that should, theoretically, replace the support team.

I wish for a shift in power: The user should be able to decide if he wants such features. Currently, you might be able to disable quick search or image AI tools on a Google Pixel smartphone, yet you’re not able to disable similar features in apps or any other place. One is at the mercy of all these companies that try to profit by implementing any such AI feature that no customer ever has asked for.

OS or browser-level AI features are just the tip of the iceberg. There are so many other features I wish I’d be able to deactivate on my devices. For example, I’ve got an iPad that is only used at home. It never leaves home, and it won’t, ever. Yet, after every update, I get a notification about setting up Apple Wallet¹. I do not want this feature. I don’t want the iPad to bother me with it either. If I want to use Apple Wallet, I’ll set it up myself, I don’t need any help with it. Another example being push notifications about using the MS store on windows machines. I don’t need and don’t want the store. I install software from wherever I want and just want to be able to disable the whole feature set with ease.

I think in general, way too many companies try to profit from AI features and bloat their products with useless stuff. This reminds me of the old days, when the browser toolbars were a place for bloatware . It’s sadly to hard for users to take action against this. Most of the time, it’s even impossible to disable unwanted features. Just think about mobile apps. The best way is to change app or provider to non-bloated (software) solutions. But in a market with only a few providers, such as the smartphone market, this is almost impossible.