If you are so keen to know, then you will just have to wait a few more years. Firefoxes development is rapidly derailing into nonsense recently. They will have to either kick out their current leadership or they will be reduced to a data sucking, adware company sooner or later.
Oh yes, as opposed to Google or Microsoft who definitely aren't already data-sucking, predatory adware companies. No thanks, I'll stick with the lesser evil.
If you're going to lie to everyone at least make it sound believable.
Do you read any tech news? If so how did u miss every single mozilla headline of the past months? Something being the lesser evil doesnt turn truths into lies.
At least this is opt-in, and Firefox still allows for manifest v3 extensions, and, on the whole, isn't using a engine funded by a billion dollar company that's doing everything in it's power to spy on you.
Yeah i was kinda overreacting but it really isnt looking good for firefoxes future at this point imo. As long as its open source there will at least be forks like librewolf.
Yeah idk either sadly. But i know that having only two relevant browsers on the market is like the US party system. Destined to fail.
Nothing lasts forever just like Steam or anything else will one day turn to shit. But pretending like everything is fine will just lead to lots of "we shouldve seen it coming".
"Trustworthy AI" + Recent aquisiton of an advertising analytics company + a call for people to inform on third party sources of Firefox = Down the enshitification rabbit hole we go.
I use a fork from F-Droid called Fennec. I'm not sure off the top of my head how closely it tracks with upstream feature-wise but I know it strips out all of Mozilla's tracking components and it's always updated within a couple days of the upstream release.
How long does AI need to be used, and how much demand needs to be sustained, for it to stop being called a "buzzword"? I'm a little dubious that NVIDIA became literally the most highly-valued company on Earth off the back of a mere "buzzword."
AI may have its uses, but the easy counterpoint to your argument is to look at FTX at its peak and where it is now (bankrupt). The stock exchange is the exact opposite of rational, and is terrible at estimating the use one can get out of tech.
Can you reminds us what the current state of NFTs is? Or most crypto? Web3 tech? This is next.
Of course Nvidia are the highest-valued company. They capitalized on idiots misusing the technology, until it created issues in society, for personal gain.
Crypto is doing kind-of ok. But what about other blockchain apps and startups, or blockchain integrations into every tech imaginable? There were so many popping up, just like there are with AI now. Business models and use-cases that are based solely on the hype of the tech in question, without any consideration about whether it's actually a good fit for the tech. That is the point, and what it has common with AI and other "buzzwords".
How do any of those things have anything to do with LLMs? You’re just listing a bunch of random tech that isn’t particularly impactful and claiming that another unrelated thing must be a failure.
Why are you explicitly picking those examples, and not things like IoT, DevOps and Edge computing, all buzzwords, all successful and still in general existence today?
You’re cherry picking failed buzzwords and using them as proof that “AI” will fail.
To be clear, I agree that LLMs are bullshit for 95% of applications they are being put into. But at least argue in good faith.
I chose those examples, because that's what's been heavily marketed recently, and it all either fundamentally failed, ended up being a scam, or both.
In contrast:
devops is software automation practices...?
edge computing is on-call load balancing? It's horrendously expensive though, so i'll give them time to figure it out
IoT, admittedly, is largely oversold, but even then, there were a ton of products on the market that absolutely outlived all 3 of the examples i've given, combined. HomeAssistant+Zigbee home automation is awesome. A raspberryPi is "iot". Your smartwatch is "iot".
There's a difference between cherry-picking, and refusing to accept that something is a scam. Crypto ended up begging for government regulation, when the original intention was to move away from it. NFTs are a pump-and-dump ponzi scheme. web3 literally doesn't mean anything
LLMs aren’t a scam, I don’t even understand how you could twist it into such. While something like NFTs have no real legitimate use case, LLMs excel at translation and as an advanced form of spelling and grammar checking.
Your complaint seems to boil down to “it doesn’t work in all use cases it’s being used” which is fair enough, but if I put a car on my bed and try to use it as a blanket… does that make it a scam?
We literally agree with each other, and yet you're still arguing. The reason why it's a scam, is because people sell it like some kind of a godsend, when it's literally not used in the way it is intended to be used. When it is, that's great. When it's trained properly, that's even better. But that's not the reality
I made a generalization based on the abundance of comments from people saying they don't want AI. Your desires may not be the desires of the majority of users.
Or maybe it's just a common fallacy. Like argumentum ad populum.
It's not. Saying a bunch of people don't want something because a bunch of people are saying they don't want it isn't argumentum ad populum. I never made an assessment about whether AI was good or bad.
If you want to argue that Lemmy doesn't represent users at large, or that the people complaining about AI are a loud minority, go for it. But the vast majority of comments on anything AI related seem opposed to it.
Using the comments from Lemmy is clearly a case of selection bias. It would be like running a poll at a gym to see how many people think exercise is important. Or asking lemmy users if Linux is better than Windows. “The people I hang around have the same opinion as me” isn’t really a good litmus test for “does this actually represent public opinion.”
I’m with you on this one. I love Lemmy, but it’s a small community here and skews towards a very specific foss tech nerd demographic that doesn’t represent the general population in any way. It seems like most users are aware of that but not everybody is self-aware enough to realize that. I like trying out AI features, I like to see them be integrated into software so they can be more useful. They’re not perfect at all but just because they’re not perfect doesn’t mean they should be abandoned in their entirety.
Can I ask areal question? I’m not trying to be a dick or smart ass, I legit don’t get this. What is bullshit here? I read the article and it seems like a useful feature to me.
“this week, we will launch an opt-in experiment”
“those who have opted-in will have the option to access their preferred AI service from the Firefox sidebar”
Is this opt in only feature really terrible? Because as a user of ai, not switching tabs sounds like a nice new feature to me.
I strongly believe that generative AI is catastrophically misused in the vast majority of its applications, so in my eyes, adding gpt-based AI to the browser is largely a wasted effort
I highly doubt they have one team that switches between experiments and bug fixes, never doing two things at once. Not to mention that something ultimately being ripped out isn’t necessarily wasted effort. They could likely easily pivot virtually anything they put into this specific experiment into any number of other uses.
I think it makes sense. I like ChatGPT and I appreciate having easy access to it. What I really wish is the option to use local models instead. I realize most people don't have machines that can tokenize quickly enough but for those that do...
Whether it’s a local or a cloud-based model, if you want to use AI, we think you should have the freedom to use (or not use) the tools that best suit your needs
They're not just giving these AI companies your data...
It's an optional feature, and you would choose which model you use. If you choose not to use it, or disable the feature, nobody will recieve your data. If you want a browser without these features, Librewolf will likely be a safe choice, as I don't seem them adding this.
The only active AI feature is the automatic alt text one, and that's entirely local. The second one is a sidebar that will just open AI chat websites, which you could already do by just, ya know, looking up the webpages the regular way. No data is getting sent anywhere so far.
people please actually read the article not the headline; this is literally about accessibility improvements for blind and visually impaired people for generating alt text inside of documents and pdfs.
people please actually read the article not the headline; this is literally about accessibility improvements for blind and visually impaired people for generating alt text inside of documents and pdfs.
It doesn't just read the page to them, which is a solved problem, it generates descriptions when they're missing, making the web more accessible.
Just curious, how do you translate things? I know Mozilla recently did some local translation stuff in-browser, but what about before? Is there a good competitor to Google Translate?
Many of the people complaining about a feature they would just disable and never use are also the same kinds of people who would complain about basic accessibility features and call them “unnecessary bloat”.
That's one of the things, but it's also adding a dedicated sidebar for AI. That's the sort of thing that should just be an extension, there's absolutely no reason at all why that needs to be something built into the browser.
Developers should be providing alt text themselves, but in cases where they aren't having a local image recognition model running to provide a description isn't terrible as long as it's either 100% local or completely opt-in.
The dedicated sidebar on the other hand feels very much like a cheap attempt to cash in on the AI fad.
That’s the sort of thing that should just be an extension
It most likely is on the technical level, just shipped by default and integrated into standard settings instead of the add-on ones. And it's going to be opt-in, so you won't have to go into about:config to disable it. Speaking of: You're looking for extensions.pocket.enabled, it should be false. And before you say "muh diskspace" it's probably like 5k of js and css or such.
access their preferred AI service from the Firefox sidebar to summarize information, simplify language, or test their knowledge, all without leaving their current web page.
Our initial offering will include ChatGPT, Google Gemini, HuggingChat, and Le Chat Mistral
Mixtral is the best of all without doubts. I use as a daily driver too and thr fact that it has loose censorship (until now) give me always great awnsers.