We’ve made the hard decision to end our experiment with Mozilla.social and will shut down the Mastodon instance on December 17, 2024. Thank you for being part of the Mozilla.social community and providing feedback during our closed beta. You can continue to use Mozilla.social until December 17. Before that date, you can download your data here (https://mozilla.social/settings/export), and migrate your account to another instance following these instructions (https://support.mozilla.org/en-US/kb/mozilla-social-faq).
Think like Microsoft recall for your browser history. Yeah that sounds awful, it’s not surprising why people would not want it.
Just goes to show that Mozilla is dumb and blind like all other companies.
So you can find things by “that spicy chicken recipe” instead of having to remember what it was actually called, or slog through a gazillion chicken recipes in your history when you realise that “spicy” was nowhere in the name. Basically stemming/thesaurus search on steroids.
It’s quite likely to be opt-in as I imagine ingesting the sites you’re looking at is a significant computational load. The translators are also opt-in, there’s enough stuff inbuilt to detect languages but not to translate, you have to download those models first. And they’re quite good btw.
Another thing I could see them offering is stuff like tl;dr bot. It’s probably not for everyone, but I definitely can see that it can be a useful feature for many people.
What do you mean ingesting browsing history locally? For AI? I don’t want that. But I’d interested in your opinion and explanation though!
Think like Microsoft recall for your browser history. Yeah that sounds awful, it’s not surprising why people would not want it. Just goes to show that Mozilla is dumb and blind like all other companies.
Stuff like Recall is okay if it’s local and opt-in
Yea…sounds terrible. I don’t even save my browser history normally.
So you can find things by “that spicy chicken recipe” instead of having to remember what it was actually called, or slog through a gazillion chicken recipes in your history when you realise that “spicy” was nowhere in the name. Basically stemming/thesaurus search on steroids.
It’s quite likely to be opt-in as I imagine ingesting the sites you’re looking at is a significant computational load. The translators are also opt-in, there’s enough stuff inbuilt to detect languages but not to translate, you have to download those models first. And they’re quite good btw.
Another thing I could see them offering is stuff like tl;dr bot. It’s probably not for everyone, but I definitely can see that it can be a useful feature for many people.
If it’s all local and no telemetry is sent to mozilla or 3rd parties, then I’d potentially be interested in it.