Just found a “
Summarize” button at the end of one thread, and it would make discourse generate an AI summary for the topic if clicked…
None of other Discourse forums I visit have this enabled.
1 Like
It is very handy for threads with lots of technical details, to get an idea or a second opinion quickly on what’s going on.
unfortunately, for large threads (like the H265 one) this option is missing.
How can you be sure what it represents is the right idea?
If someone do not understand what’s going on, they could just asked, right? That’s what forums are for.
1 Like
obviously we cannot, but it helps to get a quick overview, helped me in some complicated discussions.
1 Like
AI significantly simplifies and speeds up some of my tasks. Yes, it can make mistakes, but I’m conscious enough to notice them. Why should I deprive myself from this? Why spend an hour on something I can do in a minute?
I don’t like people who creates brainrot shorts, wrong guides, or talking with it about their people relationships, health etc. which are topics that AI cannot handle.
Nor do I like people like you who complain about everything that involves AI. Like the Firefox incident. It’s optional but everyone goes crazy about this. If you don’t like it just close it. How do you expect Firefox to compete with other browsers and appeal to the end user? End users likes AI and almost all other browsers have integrated AI. Moreover, they do this locally, in a way that’s different from other browsers. What’s the problem?
You’re all acting like a sheep. Please use your own brain and have your own opinions. Not the other ones.
Good for you. Not everyone can do this though.
Who said anyone’s going to do so? But why’s it pushed to everyone instead of being an optional feature that is disabled by default?
Disable it by default and let people enable it themselves, then. If it requires people to mess up with about:config knobs to disable it then people not extremely familiar with firefox are destined to have to leave it on and be harassed by notifications promoting AI features. That is, not yet speaking of the actual potential harm of AI features.
You clearly don’t have basic respect for people’s opinion. How about you stop replying and bragging about how good you’re at spotting bullshit generated by LLMs, and instead leave me be?
2 Likes
Cool; I’m into a posting style AI can’t understand 
I don’t like the idea of AI giving a different presentation of what I typed; I can’t guarantee how AI handles nuance or sarcasm.
2 Likes
Best letting sarcasm beside. It causes confusion an a lot of work for other users which try to manage this Discourse instance.
I guess you know what I am talking about.
And yes I still prefer to talk direct with human beings instead of AI.
Lately I made a request because of a product description in a online store. AI answered me, and was reading me the wrong product description I wanted to questioning 
I probably still would talk with AI there if i not had chosen, talk with the sales staff in person.
How do you expect Firefox to compete with other browsers and appeal to the end user? End users likes AI and almost all other browsers have integrated AI. Moreover, they do this locally, in a way that’s different from other browsers. What’s the problem?
The problem is two folds: first, people don’t “want” things they don’t even know they exist, they are convinced they need those things with marketing. AI (which BTW is not a real AI, fortunately) has being aggressivily promoted 24/7 for a while and that is like the “jedi mind trick” but on an industrial scale. It is not done because whoever pays for everything loves humanity, it is done because it means pupulation control and power (of course, money too).
The second part is “do Mozilla (or any other entity) needs to give people what people want?”. if the answer is “yes” then I would expect Mozilla to quit all the speeches about ethics and values because to align with what people wants you need to aid with the worst of humanity in any possible direction because, like the ancients said, people want “bread and games” (where “games” means people being tortured and killed). One of these directions nowadays and since ever is dictatorship and suppression of dissent, with all the consequences.
I do not use any “AI” and I don’t need it.
I find sadly amusing that people don’t undestand that “AI” is just another round of automation and that is going to make lots of desk jobs useless much like it has already happened with home banking, ecommerce and so on. At this point I see humanity in two categories, those few who will benefit from the changes and will live in fortified cities “on top of the hill” and the majority of slaves who will live in a worldwide ghetto, depending from the machines and services for living. The surreal thing is the slaves are conditioned into thinking they are “free” and whoever controls them is “the big brother”, loves them, provides for them. So they are grateful.
I would like to write more about some “geopolitics” connected to IT and services but I would probably get censored.
2 Likes
Does it have any sense to discuss this?
the button is completely non-intrusive and optional.
Whoever wants the AI generated summary and finds it helpful, can click it at one’s own will.
The rest just don’t click it.
It’s just a button…
1 Like
First, enabled by default and allowing switching off is “opt-out”, not “opt-in”.
And you know how most users click through settings and never review them again.
Second, attacking people’s (not really) out of the latest news cycle by nationality is pretty shitty behavior and will be reported to the moderator. (Do you think people in China get to opt-out of that GFW thing, btw?)
Third, I believe I already kindly asked you to leave me alone.
2 Likes
If you can connect:
This isn’t your home, this is a forum. You created a discussion we’re discussing. If you don’t want to discuss, you can leave the forum; nobody’s stopping you.
@libpaspivipso, I agree that it can be annoying to see AI tools everywhere. We have to differenciate between hype and usefulness though.
There are some really long topics here on Fedora Discussion (e.g. this, or this one), where someone not being involved in the discussion threads from the beginning might find summaries helpful.
While adjacent to this topic, it’s worth mentioning that there was a community review and feedback collection this year on a council policy proposal regarding AI-assisted contributions.
I believe I have made myself abundantly clear that “default on but allow users to turn it off” (opt-out) != “default off” (opt-in) because many/most users don’t change their default settings. The difference being they didn’t ask the users before turning it on.
If they offer this kill-switch on browser first run setup (something like "do you want to switch all AI features in Firefox off?) then it’s probably less bad, but still worse than ask before turning it on (“do you want to enable AI features in Firefox?”).
(Also yes I can open that web page without violating the law, thanks for asking)
The real reason the AI summary exists is because Discourse’s infinite scrolling means that it is impossible to reliably consume the whole discussion. With long threads, you basically have to give up because the lack of paging means you can’t do normal progression of the thread.
Discourse is a horrible system for long discussions because of it.
5 Likes
I am not an “AI” fan, far from it, in fact. But the “Summarize” function in Discourse is one area where an LLM (Large Language Model, which is what most people refer to when they use “AI”) actually makes sense: it takes some text and produces some other (shorter) text. Can anyone guarantee that it is without error? No. But the same is true for any encyclopedia, like Wikipedia. If you want to be sure, you have to read the source material (i.e., the original post(s)).
And it is a button, nobody has to click it. In fact, I would go so far as to say that pretty much every software I use has some amount of functionality I do not need or use. But that doesn’t mean I have a right to complain about it and demand its removal. (Well, I guess I could do that, old man yelling at clouds, etc., but it doesn’t feel particularly useful.)
2 Likes
I do not know if you heard about the /raw/ version of a topic? I guess so, however others are also reading here … I do not know what would be ideal for you? if git oriented content would be that what you are looking for?
3 Likes
It’s something, but I would rather have static pagination like every other forum solution.
1 Like
Didn’t know that, thank you. That’s pretty interesting a feature.
1 Like