After trying the app, it's hard for me to interpret this article as anything other than Mozilla lying. Sharing in this app is the same as any other social media app.
In the app there is a "Share" button at the top right. After clicking you see an interstitial with a big "Post" button at the bottom. When you click that button, the chat is shared.
Am I seeing something different than anybody else? Why would Mozilla lie like this? Most of the "demands" are already satisfied.
> Shut down the Discover feed until real privacy protections are in place.
Everything is already private by default and you can see what is public.
> Make all AI interactions private by default with no public sharing option unless explicitly enabled through informed consent.
This is true already
> Provide full transparency about how many users have unknowingly shared private information.
Meta shouldn't have to do this
> Create a universal, easy-to-use opt-out system for all Meta platforms that prevents user data from being used for AI training.
This already exists (EDIT, looks like only for EU users. Personally I don't believe this is related to the public sharing claims)
> Notify all users whose conversations may have been made public, and allow them to delete their content permanently.
This is a dark-pattern problem. A large number of people are accidentally sharing things to the general public when they intended to share them to specific people. That is one issue being flagged here. To many people, "share" means "give me a way to share to specific people", not "mark this for indexing/searching for the general public".
The difference is that on those apps it would be a miracle if your "publicly" shared post were ever seen by more than a handful of strangers. None of those send it to the front page except in very rare circumstances (going viral).
The point is not the fanout factor but the lack of limitations. A public post is such that anyone can see, without authentication. A widely shared post that requires login to see is not public. An obscure pastebin paste is public, even though it's normally seen by a handful. But you don't get to control who these handful are.
But it is different? In both Gemini and ChatGPT when you click "Share" you get a link to the post you can share. It doesn't add the chat to common "Discover" section in the app (there is no such section there). As others pointed out "Share" in Meta AI app is actually "Publish", unlike the one in other chat apps where it is, in fact, share.
The action button says "Post", which to me is pretty close to Publish
And shared ChatGPT chats are often indexed by Google, so they become public. Although I agree it's not exactly the same because of the lack of builtin discovery
“Post” is indicative for a person paying attention. For a random non tech person trying to share thing it’s just one more button to press to get the link. Note it’s very hard to share the chat while not making it public - you will need to explicitly mark it private after you posted it. I bet you for absolutely majority of users posts being public and discoverable pretty much unintended consequence of them trying to share it to someone.
As for google I assume it’s true if you post your chat somewhere (I.e Reddit). For most of shares these chats never end up being on the internet (I.e stay in private chats in messengers). So it is different again.
Overall it is pretty much predator behavior exploiting people’s need to share their chats to get them doing something unintended and good for Meta. This being said whole idea of an app with chats as posts is quite lame, so not sure if it will stick.
>And shared ChatGPT chats are often indexed by Google, so they become public. Although I agree it's not exactly the same because of the lack of builtin discovery
I didn't know this. I think I have used share button to share chats with personal stuffs with close people thinking that the chat is as close as possible to "unlisted" youtube videos. Hope that OpenAI implements something similar or even something with a password.
> And shared ChatGPT chats are often indexed by Google, so they become public.
Do you mean creating the sharing link would make the chat log public? It will be interesting to read other people's chat on a specific topic, but googling "topic site:chatgpt.com" seems to just return Custom GPT stuff.
What is the purpose of quoting "topic" as if it is a search keyword? Are you searching for this literal string in the logs? Why are you bundling it with the "site:" operator?
I got an empty search results error with your suggested turd.
I don't think this is a dark-pattern problem in the sense that I don't think it is _intentionally_ deceiving.
I think Meta fully expected this feature to be used by people who are excited about their conversation with the AI and wants to share it publicly. Just like we see with OpenAI Sora.
There's not much to win for Meta if users instead are unknowingly sharing deeply personal conversations.
> I think Meta fully expected this feature to be used by people who are excited about their conversation with the AI and wants to share it publicly.
That's really what you think? And what they think? That people are so enamored - in droves - with their exchange with a chatbot that they're trying to share it for the world to see?
Maybe I'm the old fogey who doesn't get it, but it's just hard for me to believe that this is something many people want, or something that smart people think others earnestly want. Again, I may be the outlier here, but this just sounds crazy to me.
People share AI chats all the time on Twitter, Reddit, etc.
I don't personally think the feature makes a lot of sense in Meta AI.
However it's a lot more likely their product team genuinely thought it might do, than it is likely they intentionally wanted to give users a bad experience and risk more bad press (again, Meta would benefit nothing from people sharing by mistake).
Considering that 90% of the chats I see share are people tripping over themselves to demonstrate the AI being silly and dumb, yeah they are enamored to share with the public :p
I agree. Further, these companies show us over and over again who they are, and whether it's tobacco companies, pharma, food, or oil companies they always know - in exactly the way and at the time that makes you sick to your stomach - what they're doing and who's likely to fall for it in a way that makes. The comments in this topic are feeling a bit sophist
You completely misunderstood me. I am not incredulous that people use AI, nor am I in any way doubting how it can aid all sorts of processes.
I am incredulous that a primary use case of a genAI chatbot is sharing your chat conversation publicly. It's easy to see why people would do this for genAI images, videos, or even code; I even understand some occasional sharing of a chat exchange from time to time. But routine, regular interest, from regular people, of just sharing their text chat? I do not understand that at all.
> I think Meta fully expected this feature to be used by people who are excited about their conversation with the AI and wants to share it publicly. Just like we see with OpenAI Sora.
META expectations=/= expectations of a reasonable human that has used other "share" buttons before.
Sharing to a text message is private. In contrast, sharing to social media platforms such as Twitter, Reddit, Pinterest, and LinkedIn makes the content public. The destination determines the audience.
The TYPICAL behaviour when you hit "share" on any platform is not to immediately share. It TYPICALLY gives you options to share to a variety of other sources, both public and private. It also generates a link if you want to grab the link and specifically share that.
That is the TYPICAL share behaviour. If what META is doing with their new app is obscuring this typical behaviour and a "share" click directly going to the public, that would violate the defacto behaviour users are accustomed to when using the share button.
That is possible. I wouldn't think that because there are no "friends" in this app but I could see why a Facebook user might think that. On the other hand, when you open the app you immediately see content from people you aren't connected to. It all feels very public to me.
I opened the app and the third post was someone making a note to self to cancel their car insurance, followed by a reply comment saying oops that wasn't supposed to be public, so at least one user was confused.
Your question is important because we need to understand nothing is private online. Yes, thankfully our bank accounts and other important info is PW protected, however, these PW's are eventually stolen by data breaches. (Didn't we all recently have to change our PW's on FB, Microsoft, Google and Apple?)
To think that anything used on AI is going to stay private is nice, but not likely.
> Am I seeing something different than anybody else?
Maybe. Maybe today, maybe tomorrow.
As others have mentioned, the core problem with Meta today is the dark patterns. They move, edit, and remove UI elements specifically to optimize against whatever behavior they want the user to take. I'm always amazed when things end up posted, shared, or alterated in a way I did not intened or can't even remember having taken an action against. Things just seem to happen with Meta products… even for accounts that are idle.
And if you spend enough time with Meta products, you'll start to realize that no two users are guaranteed to have the same experience. There is no standard experience. The experience changes based on region and langauge and honestly who knows what else. They are constantly testing and optimizing for dark patterns in production. Spend an hour with the Meta Business Suite. The entire platform is essentially a dark pattern labyrinth of broken links, broken features, and UI elements that go nowhere or to deprecated functions. One team is trying to get you do X and use feature Y, and another team is trying to get you to do Z and use feature W. Business Suite just mashes it all together. You could freeze the codebase today and study Business Suite for months and you'd find that it's dark patterns all the way down.
Users are playing around with AIs for entertainment all the time. You wouldn't be able to determine if seemingly private information was real or made up.
In the app there is a "Share" button at the top right. After clicking you see an interstitial with a big "Post" button at the bottom. When you click that button, the chat is shared.
Am I seeing something different than anybody else? Why would Mozilla lie like this? Most of the "demands" are already satisfied.
> Shut down the Discover feed until real privacy protections are in place.
Everything is already private by default and you can see what is public.
> Make all AI interactions private by default with no public sharing option unless explicitly enabled through informed consent.
This is true already
> Provide full transparency about how many users have unknowingly shared private information.
Meta shouldn't have to do this
> Create a universal, easy-to-use opt-out system for all Meta platforms that prevents user data from being used for AI training.
This already exists (EDIT, looks like only for EU users. Personally I don't believe this is related to the public sharing claims)
> Notify all users whose conversations may have been made public, and allow them to delete their content permanently.
This already exists