Google could be about to let you filter Gemini output
long story short
- Google may be about to give users new settings to control how Gemini reacts.
- This “content filter” may control the level of censorship in Gemini itself.
- The Gemini API already provides developers with “hazard category” settings.
Last year, artificial intelligence has penetrated almost every field. When it comes to artificial intelligence and Google, Gemini is the name of the game. The company has been working on improving its model and announced earlier this month Gemini 2.0. Because of the power that AI offers, companies like Google spend a lot of time thinking about what kind of protections their AI systems need to have in place. We’re currently working on a new way Gemini may soon give users more control over the type of content they generate.
one APK teardown Helps predict future features that may appear on the service based on ongoing coding. However, such predicted functionality may not be publicly released.
Today we’re taking a look at the 15.51.24.sa.arm64 beta version of the Google app for Android. While this isn’t public and live yet, we were able to enable a new option under Gemini settings that appears to be preparing the ability to filter Gemini output by content.
AssembleDebug / Android Authority
Currently, selecting this option does nothing useful and just takes us to a URL on the Google Gemini website that has not yet been launched. While this makes it difficult to determine how this setup works, we can make some inferences based on what we know about Gemini.
Google has some formal policies in place What can you do with Gemini?and by extension, also has some goals for the type of content it is trying to achieve Guide Gemini to avoid. In particular developers have access to some fairly fine-grained Gemini security settingsgiving them control over a number of “categories of harm,” including hate speech, harassment and sexually explicit content.
After seeing all the features currently available, Google may extend some level of the same control to Gemini end users on Android and the web. That said, Google could go in any direction when it comes to the implementation here, from the broad, developer-level filters that the API now has, to potentially more restrictive filters, all to prevent Gemini avoids sharing anything particularly offensive.
At this time, it’s difficult to say when this new option will go live, but we’ll be keeping an eye out for any further developments.
2024-12-26 22:26:20