I mean, they’ve been all over Reddit running accounts for decades at this point, no? At this point I feel like 75% of my interactions on mainstream sites are going to be from foreign agents and/or bots.
They simply ban you there for a mere criticism of Israel, which has nothing to do with their T&C. By now this subreddit turned into an echo chamber of likely minded bots and trolls.
That was an annoying read. It doesn’t say what this actually is.
It’s not a new LLM. Chat with RTX is specifically software to do inference (=use LLMs) at home, while using the hardware acceleration of RTX cards. There are several projects that do this, though they might not be quite as optimized for NVIDIA’s hardware.
Go directly to NVIDIA to avoid the clickbait.
Chat with RTX uses retrieval-augmented generation (RAG), NVIDIA TensorRT-LLM software and NVIDIA RTX acceleration to bring generative AI capabilities to local, GeForce-powered Windows PCs. Users can quickly, easily connect local files on a PC as a dataset to an open-source large language model like Mistral or Llama 2, enabling queries for quick, contextually relevant answers.
Pretty much every LLM you can download already has CUDA support via PyTorch.
However, some of the easier to use frontends don’t use GPU acceleration because it’s a bit of a pain to configure across a wide range of hardware models and driver versions. IIRC GPT4All does not use GPU acceleration yet (might need outdated; I haven’t checked in a while).
If this makes local LLMs more accessible to people who are not familiar with setting up a CUDA development environment or Python venvs, that’s great news.
Ooh nice. Looking at the change logs, looks like they added Vulkan acceleration back in September. Probably not as good as CUDA/Metal on supported hardware though.
How has no one worked on a new search engine over the last decade or so where Google has been on a clear decline in its flagship product!
I know of the likes of DDG, and Bing has worked hard to catch up, but I’m genuinely surprised that a startup hasn’t risen to find a novel way of attacking reliable web search. Some will say it’s a “solved problem”, but I’d argue that it was, but no longer.
A web search engine that crawls and searches historic versions of a web page could be an incredibly useful resource. If someone can also find a novel way to rank and crawl web applications or to find ways to “open” the closed web, it could pair with web search to be a genuine Google killer.
There’s a lot of startups trying to make better search engines. Brave for example is one of them. There’s even one Lemmy user, but I forget what the name of theirs is.
But it’s borderline impossible. In the old days, Google used webscrapers and key word search. When people started uploading the whole dictionary in white text on their pages, Google added some antispam and context logic. When that got beat, they handled web credibility by the number of “inlinks” from other websites. Then SEO came out to beat link farmers, and you know the rest from there.
An indexable version of Archive.org is feasible, borderline trivial with ElasticSearch, but the problem is who wants that? Sure you want I may, but no one else cares. Also, let’s say you want to search up something specific - each page could be indexed, with slight differences, thousands of times. Which one will you pick? Maybe you’ll want to set your “search date” to a specific year? Well guess what, Google has that feature as well.
They’ve had a history of controversy over their life, ranging from replacing ads with their own affiliate links to bundling an opt-out crypto miner. Every time something like this happened, the CEO went on a marketing campaign across social media, effectively drowning out the controversial story with an influx of new users. The CEO meanwhile has got in trouble for his comments on same-sex marriage and covid-19.
In general, it’s always seemed like it would take a very small sack of money for Brave to sell out its users. Also, their browser is Chromium based, so it’s still contributing to Google’s market dominance and dictatorial position over web technologies.
Bing’s copilot is genuinely pretty good, the AI answer is often pretty accurate and the way it’s able to weave links into its answer is handy. I find it way more useful than Google search these days and I’m pretty much just using it on principle as Google is just pissing me off with killing their services, a few of which I’ve used.
I don’t think Microsoft is some saint but copilot is just a good product.
Yes, that would be a Google killer. If you somehow find the money to provide it for free.
Finding a novel way of searching is one thing. Finding a novel way of financing the whole endeavor (and not going the exact route Google is) is another.
This feels a bit like asking MS Teams to play nice with Google Meet, or demanding that Apple’s office suite (Pages, Numbers, etc.) deliver the exact same product when files are saved in an OpenOffice format. This doesn’t seem to be an issue with any other products…
Apple have designed their product to work well with their devices. The Messages app still functions with non-Apple devices. SMS messages can be sent and received to anyone. The fact that pictures and whatever come through like crap is more an issue with the SMS platform than it is with Apple’s app.
Ultimately, Google dislikes the fact that there is a “green bubble” stigma (for lack of a better word) on Apple devices that encourages those who care about such things to prefer Apple devices. Because Google doesn’t have their own widely used iMessage equivalent, they can’t turn around and make messages outside their platform appear as red bubbles or something, so they are attacking from this angle instead.
Google are attempting to start this fight in Europe in hopes that they can push Apple to change in the US as well. The whole green bubble thing is US-only, but Google haven’t been even remotely successful in trying to force Apple to change, and Apple’s “remedy” to the issue is “Get an iPhone”.
Anybody remember Hangouts? Google’s iMessage that was better in every conceivable way than its Apple analog, integrated with Google voice, could be accessed anywhere you could get on Gmail etc? Dropping the ball on Hangouts to favor carrier pre-installed messaging Apps was such an incredible and short-sighted blunder. I concede that exactly like their many app deprecations/cut-and-runs that did not take the long-term sentiment of the end user into account and damaged their reputation and adoption. And now here we are… trying to grovel back into iMessage’s purview.
I think you’re right. I got some people to start using Hangouts and then Google killed it. I don’t even bother to learn what Google has available now for chats because I know now there’s no point to trying to get people to switch, no matter how good/bad it may be.
Wait you thought hangouts was good? Holy shit would that be one of the worst Google offerings of the decade if it wasn’t for the ten other Google chat and video systems they have made. My god I can’t think of a worse communication platform than hangouts. You might be the first person I’ve heard of liking it.
I don’t want to make assumptions, but your reply makes me think you arrived at Hangouts once it was already being deprecated by Google. Granted, being US based I didn’t need the coverage of WhatsApp (limited as they was even then to phone # accounts), the scant usage of Viber or the other innumerable messaging apps I touched in that time period. Hangouts integrated seamlessly with SMS, let me send media/stickers/map embeds to mixed-platform groups never worrying about quality downgrade. And did I mention that one could access Hangouts (and its SMS pass through server) from any machine in the world through Gmail? iMessage makes you jump hoops to do that shit today.
And did I mention that one could access Hangouts (and its SMS pass through server) from any machine in the world through Gmail?
this was actually one of the things I hated the most about it. It doesn’t really matter what features you provide when the product is so bad it can’t even make up for it. I had no clue it had sms passthrough, it was just a shitty chat/voice client integrated into my email client, slowly making things slower and slower the more they added.
Bingo. This whole case is designed to make Apple look like the bad guy whilst Google hides their real agenda of forcing Apple to use a protocol Google controls and thus stamp out Google’s competition.
From what I’ve read, Google just owns the reference implementation. Apple could implement it themselves, but then lose out on certain non-cross-platform features, like e2e encryption.
I’ve read the specification. Google’s implementation is the only real implementation (raw RCS is basically a dead project) as Google have added a load of custom extensions to RCS that means, to be interoperable, you’d need to use Google’s (which I imagine requires licensing since it doesn’t appear to be open source).
It’s RCS not RMS and Google didnt even want control of it in the first place, it’s well documented Google has been trying to get US carriers to stop dragging their feet on RCS for a long time. They never did until Google literally went “Fine, I’ll do it myself then”
AND RCS is an open protocol, nobody really has “control” over it, Google runs some RCS servers but if it disappeared tomorrow (Or you changed the defaults) RCS itself would run just fine on whatever including if Apple supports RCS
ETA: Also SMS is absolute trash, it’s from the early 90’s (it’s older than me FFS) it doesn’t really support what we want out of it media wise today, and what it does support it was forced to. It’ll send “video” but it’ll be completely unrecognizable. It needs to be put to pasture already.
It will be an easier sell if Google manages to get their proprietary extensions to RCS into RCS version 10, rather than only being supported in Google Messenger
SMS takes less bandwidth and is perfect for large broadcast messages and works perfectly fine for text based messaging. The only major problems it has are security and media, which while are valid needs, are not a reason to get rid of one of the few universally accepted standards
It doesn’t need to be a fallback. It’s still perfect for text messages, government alerts, mass notification of customers, etc.
It’s barely used today anyways. The only time it’s used on iPhone is if you’re messaging someone outside the iMessage ecosystem, which really isn’t a problem for 95% of Apple users.
I don’t know why you insist on holding onto a 30+ year old protocol. It’s not perfect and at times it can be downright unreliable. Once it’s left your phone you have no idea if it was successfully delivered or not, there’s no acknowledgement no retrys no retransmits. It just shoots it off and hopes for the best.
Group chats are laughably broken even among all SMS recipients (It was never intended for it anyway) and frankly the bandwidth required for text regardless of if it’s over SMS or RCS is inconsequential, who cares if RCS messages need a bit more bandwidth to send text. The difference is negligible.
That’s why it works so well. What you see as problems with SMS I see as good design decisions. It’s an incredibly simple implementation that does exactly what it’s supposed to. You just want it to do more than it needs to.
Something will eventually replace it, but it sure as hell won’t be RCS. RCS is a defacto google standard now. Many features are locked out if you don’t use google servers. It’s not an open standard and it’s disingenuous to portray it as one.
Uh yea, because we do so much more on our phones, it might be well designed, but it was well designed for the 90s. That’s why it makes a good fallback protocol, but by no means should it be the go-to.
RCS is the replacement, it’s been the replacement for a long time in the EU. In fact, if the US carriers just implemented it when the EU did, this entire thread wouldn’t even exist.
It’s a standard until Google takes control of the GSM Association.
The Verge in 2019 criticized the inconsistent support of RCS in the United States, with carriers not supporting RCS in all markets, not certifying service on all phones, or not yet supporting the Universal Profile. Concerns were shown over Google’s decision to run its own RCS service due to the possibility of antitrust scrutiny, but it was acknowledged that Google had to do so in order to bypass the carriers’ inconsistent support of RCS, as it wanted to have a service more comparable to Apple’s iMessage service available on Android.
You’re like a lot of people on Lemmy: so eager to paint everything even tangentially connected to Google as some kind of grand conspiracy that you can’t even get the most basic facts right.
It’s quite literally well documented that Apple doesn’t want to support RCS because it pressures people to get iPhones. SMS is an ancient garbage protocol, what Google is trying to do is get Apple to support SMSs 21st century replacement and RCS support will fix literally every issue iPhone users have texting Android users. Broken group chats, trash quality videos, ultra compressed images, no reactions or stickers, threaded chats etc etc
Google wants Apple to use Google’s proprietary extension of RCS, which runs on Google’s own servers as is precisely as open as iMessage. Effectively nobody uses the industry-standard version of it.
Where’s the source for that? Last I read, Google was using the GSMA Universal RCS profile
Google does own and run the Jibe platform as an RCS vendor, but Apple doesn’t need to use it. They can go with a different vendor or run their own RCS servers just as easily
Google's astroturf campaign for "RCS" promotes encrypted messages but RCS has no support for this. Google wants to force people to use its proprietary extension, which runs exclusively on Google's servers.
The difference is, you can choose not to use Googles RCS extension and opt for the Universal Profile standard instead and it will interop with people on other RCS profiles, even Googles, just fine.
iMessage doesn’t do any of that, your choice is iMessage with other Apple users or a 30+ year old protocol. That’s it.
Except that’s not what happened in reality before Google started rolling out their version of RCS.
The carriers implemented their own versions that didn’t weren’t interoperable with each other, and that was for the ones that even bothered with it at all.
And now they have even less incentive to try.
RCS is nice in theory, but no one is serious about implementing the universal profile.
Despite being potentially catastrophic, this issue is unlikely to impact regular people. That’s because in order to make full use of the flaw, hackers would already need deep access to an AMD-based PC or server.
Just get the service for a month and drop it after catching up on the things you want. I keep it so I can watch bluey with my daughter, but I rotate most of my services.
I think the next thing they're gonna do is go after people who sub for a month and then unsub. Probably by charging a good bit more for month-to-month than paying annually.
Might miss out on a one time deal that locks your price for a year, but even if its fifteen bucks for a month to catch up on things I want to watch I’m fine switching services each month.
So my child, who is not old enough for their own account, will now no longer be able to watch Disney+ while attending school at their residential academy 400 miles away. Just like Netflix. And just like Netflix, my subscription will be canceled the moment they try to block them from logging in.
I decided I would help Disney out with their password sharing problem by canceling my Disney Plus account a few months ago. No, they never have to worry about it again.
I’m not surprised. It went really well for Netflix.
Everybody said they’d cancel Netflix over it, even that it would be a mistake that would kill Netflix, but when it came down to it, most continued paying/bought a plan and Netflix became more profitable.
What's probably more likely is that the "everybody" that you heard from was an incredibly unrepresentative sample of people from a bubble of nerdy tech enthusiasts.
I cancelled. It’s just that I’m a small minority of people. The number that cancelled was apparently less than the people who signed up for their own account. Oh well. Netflix wasn’t that good anymore anyway. I barely used that app. Disney Plus however I’m not going to cancel. That one is worth it to me.
So my child, who is not old enough for their own account, will now no longer be able to watch Disney+ while attending school at their residential academy 400 miles away. Just like Netflix. And just like Netflix, my subscription will be canceled the moment they try to block them from logging in.
It was meant for helping people access pages when way back, you often couldn’t depend on a page loading,” Google’s Danny Sullivan wrote. “These days, things have greatly improved. So, it was decided to retire it."
They still go down, Danny. And fairly frequently at that. Y’all are fuckin’ stupid.
I’d say things are much worse than they used to be. Sure, in the past sites would disappear or completely fail more often. But, because most sites were static, those were the only ways they could fail. These days the cache feature is useful for websites that have javascript bugs preventing them from displaying properly, or where the content-management-system still pretends the link works but where it silently just loads different content.
By they way, I just found out that they removed the button, but typing cache:www.example.com into Google still redirects you to the cached version (if it exists). But who knows for how long. And there’s the question whether they’ll continue to cache new pages.
I hope they only kill the announced feature but keep the cache part.
Just today I had to use it because some random rss aggregator website had the search result I wanted but redirected me somewhere completely different…
Quotes are fucking awful now. You have to change the search terms to verbatim now which takes way fucking longer. Google has enshittified almost everything. I’m just waiting for them to ruin Maps.
My guess is that a cached page is just a byproduct when the page is indexed by the crawler. The need a local copy to parse text, links etc. and see the difference to the previous page.
engadget.com
Top