Firefox 148 introduces the promised AI kill switch for people who aren't into LLMs
Firefox 148 introduces the promised AI kill switch for people who aren't into LLMs
Mozilla has released so many self-described AI features in the past few years, but this is the only one that has:
I hope Mozilla learns their lesson. I doubt they will, but I hope.
When Firefox 118 was announced, they didn’t call it AI. They didn’t even call it machine translation, which is what it was.
They called it local, automated translation.
Maybe you should have read what I wrote and what Mozilla said.
And?
Because the term AI was not in vogue at the time, even though it’s clearly the same technology, it doesn’t count? It’s literally packaged under the same umbrella now.
Anyway, the big issue is still tech ppl thinking their viewpoint is the only one valid, and that every generic user will have the same exact needs as them.

Federated, open‑source, ad‑free, and fully under your control. Build or join a community that reflects your values with no corporate overlords. This instance is run by the founder of PieFed. [Mobile apps for PieFed](https://piefed.social/post/1258559)
You’re defending your position that this AI feature is not really AI so it’s ok
I literally say “The translation is technically AI,” so no. I give reasons how the other features are different, which you seem to acknowledge a little, at least.
the weird gimmick you don’t understand is crucial to some
Can you describe how to access the gimmick and which people find it crucial? I’m pretty confident in my understanding of it and how hilariously unhelpful it is.
Being technically something implies it’s not really or to be considered apart from the group.
The “gimmick” is proposing alt text based on the image when editing PDFs. I don’t see how it’s unhelpful. I’m not into editing PDFs in firefox, but I do use it to read them.
Inciting editors to include an alt text for accessibility seems like the ideal use case for this tech. The human still has to review and approve the generated text.
Unless I missed something as I cannot try the feature now, it seems to me a great application of ai, to augment humans in their work, and to a useful cause.
Image classification and description is “old” tech now, and I already use it in my work to auto tag images for editors to find more easily later. Nothing crazy.
The “gimmick” is proposing alt text based on the image when editing PDFs. I don’t see how it’s unhelpful.
A gambling toolbar that links to Polymarket could be helpful. But I think we both said “crucial”.
If you know someone who uses Firefox to add images to PDFs so often that the alt text generation would be crucial to them, or even more than a gimmick, please introduce me to them. I have so many burning questions. Several things related to “why not a dedicated PDF editor?!”
There enlies my point. Mozilla added a feature that is buried so deep in Firefox that you don’t even know if it’s actually there.
It is there, believe it or not. I criticize these things, but I also tested it the day it arrived.
But despite the way Mozilla buried it, the code is still there. It still makes Firefox more complex to maintain, and Mozilla still spent time and money putting it in. Imagine if Mozilla spent those resources actually trying to help people, instead of treating AI like companies used to treat blockchain: as a solution looking for a problem.