Just want to raise up the design of this “AI kill switch” in Firefox, and say how much I like it: it’s a top-level preference, and the design and content are both *very* clear in communicating the different levels at which you’re making decisions.

I would *love* to see more product companies roll out something like this for their en-LLM’d features.

In before the replies to say:

1. yes, I wish it was opt-in;
2. yes, I’d much rather something like this didn’t need to be designed in the first place

Given all that, and until that happy day that “AI” platforms are rendered down into their component pieces and sold for scrap, I think this is a really, really good piece of design.

(Thanks so much to @adarsh for flagging it in the first place: https://ruby.social/@adarsh/116138039920943577)
adarsh 🚲 (@[email protected])

Attached: 1 image Good for #Firefox for finally putting in a "Please for the love of all that is holy, no AI anything ever" button into release 148. I am *very* curious to know the stats on uptake over time.

Ruby.social
@beep i get the impression we haven't seen it more because for the companies that would be adding such a feature it is all downside: work to do, and gives them a concrete internal number of what % of their users explicitly reject this thing they've essentially bet the farm on. and even mozilla i'm guessing wouldn't have done this if tons of people hadn't yelled at them. it's gonna be an interesting year.
@jplebreton Oh agreed, the incentive structures just aren’t there during this whole gold rush. That’s maybe another reason I like this design, though? Yes, it took a lot of external pushback, *and* Mozilla could’ve easily settled on a middling design. I love where they landed!

@beep I am once again wondering about selection bias.

What portion of users who eagerly check the AI kill box will have not also unchecked the anonymous metrics opt-in box?