lloram239
@lloram239@feddit.de
- Comment on Tomb Raider I-II-III Remastered details enhancements, new features 9 months ago:
The original games are so old that it is quite messy to get them to run on PC these days and the remake changed the gameplay quite substantially. So this remaster is quite welcome and probably a lot cheaper to produce than a full 2 and 3 remake.
As for Anniversary, the story there is a bit weird, as Anniversary and Legend are basically completely independent games that have nothing much to do with each other, Legend is a reboot and Anniversary is a remake, it’s only with Underworld that the the story of those two get wrangled together into a trilogy. Only the first game is taken into account there, all the other sequels of the original game are ignored.
- Comment on Google Search Really Has Gotten Worse, Researchers Find 9 months ago:
Brave has their own index as well. And if you want Google results in not-enshittified, try Kagi.
That aside, the biggest frustrating in the search space is the complete lack of innovation. All those search engines and their alternatives do the same thing and look the same. There haven’t been new features or new sources of information in about a decade. The whole space has been extremely stagnant.
The only new thing we got recently was ChatGPT, but as search engine replacement it really doesn’t cut it right now, it can enter Wikipedia-style general knowledge question ok’ish, but completely falls apart on anything even mildly obscure (e.g. summaries of lesser known movies are completely wrong). I hope that something good comes from all the AI development, but BingChat so far is a really lack luster and ham-fisted attempt at integrating search with AI, often performing much worse than plain ChatGPT instead of better.
- Comment on Google Search Really Has Gotten Worse, Researchers Find 9 months ago:
You have plenty of literature that can act as ground truth. This is not a terribly hard problem to solve, it just requires actually focusing on it. Which so far simply hasn’t been done. ChatGPT is just the first “look, this can generate text”. It was never meant to do anything useful by itself or stick to the truth. That all still has to be developed. ChatGPT simply demonstrates that LLM can process natural language really well. It’s the first step in this, not the last.
- Comment on Google Search Really Has Gotten Worse, Researchers Find 9 months ago:
It’s not about identifying AI or even spam, but about extracting useful information. Are the claims made in a source backed by other sources? Do they violate information from trusted sources? That’s all stuff that an AI can reason about and then discard the source as junk or condense it down to the useful information in it.
Basically you completely skip browsing the Web yourself and just use the AI to find you what you want. Think of it like some IMDB or Wikipedia, but covering everything and written and curated by AI. When the AI doesn’t already know some fact, it goes crawling the Web and finding it out for you, expanding its knowledge base in the process.
At the moment there are still some technical hurdles, the AI systems we have are all still a little to stupid for this. But that seems to be the direction we are heading, things like summarizer bots already do a pretty good job and ChatGPT is reasonably good at answering basic questions. Only a matter of time until it gets good enough that you couldn’t do a better job yourself.
- Comment on This tool for finding bad USB-C cables 9 months ago:
If you wiggle them around long enough, they all go bad. But it should take a few years under normal use. They are rated for around 10,000 insertions. Cable bending can also damage them.
The more annoying part with modern USB is that not all cables are alike to begin with. Cheap charging cables that you get with random gadgets (e.g. flashlights, fans, etc.) will often just have two pins connected, meaning they work only for charging, not data. Others might have data pins, but only enough for USB2, not USB3 speeds. Others might have too much resistance slowing down charging or dropping too much voltage to even have a device function properly at the other end (common issue with long cables or extensions). And so on. Rather annoying to deal with when you just have some random cables floating around, as there is absolutely no labeling or color coding to differentiate the cables.
- Comment on Google Search Really Has Gotten Worse, Researchers Find 9 months ago:
For quite a while by now. Three years ago or so they started recommending older content again, instead of focusing exclusively on new stuff. And since than I frequently end up on videos and tiny channels with just hundreds of views. Meanwhile on a regular Google Web search I literally never end up on somebodies random private homepage, I have to remember that Marginalia exist if I ever wanna see one of those.
Youtube of course still favors professional monetized content, but random niche content still ends up making its way to the top surprisingly often.
- Comment on Google Search Really Has Gotten Worse, Researchers Find 9 months ago:
While Web search has gotten worse, Youtube has gotten pretty good at finding niche content with a few dozens views. In general it seems most user generated content these days is on Youtube as video, not on the Web as text. The typical Web SOC spam doesn’t really exist on Youtube outside of a few crypto scams here and there.
- Comment on Google Search Really Has Gotten Worse, Researchers Find 9 months ago:
I am more optimistic on that one. AI provides a pretty clear way out of this, since it allows you to automatically detect the bullshit. Meaning either the bullshit has to raise so much in quality that it is indistinguishable from good content, in which case it would not be bullshit anymore, or it will get filtered. AI can also transform bad websites into good ones, like a super-powered ReaderMode, so a lot of the “lets plaster everything with ads” will lose effectivness.
The problem over the last decade was that Google completely lost interest in being a search engine, they are just an ad company and as long as search leads you to more ads, they are quite happy. So the user experience went down the toilet.
- Comment on Google Search Really Has Gotten Worse, Researchers Find 9 months ago:
!gjust redirects you to Google search. DDG itself is just Bing with extra marketing. If you want Google+cleanup you have to use Kagi, which gathers its results by combining different sources. - Comment on AI comes up with battery design that uses 70 per cent less lithium 9 months ago:
Straight from Amazon, eBay, Aliexpress and Co. All still early days, but things are ramping up.
- Comment on AI comes up with battery design that uses 70 per cent less lithium 9 months ago:
Call me when there is an actual battery based off peer reviewed research that has been successfully tested in production systems by at least 5 major companies.
While everybody was busy writing bullshit hype articles, we actually got a real revolution with the sodium-ion battery, which you can buy today. It won’t replace Li-ion in terms of energy density, but it’s much more robust, cheap, handles low temperatures, deep discharge and much more charge cycles, making it ideal for off-grid-storage.
I really wish we had tech news that just reports on stuff that’s tested and available for purchase. Things do actually keep improving, but it’s completely drowned out in all the other hype.
- Comment on YouTube is slowing down for users with ad blockers in new wave 10 months ago:
I don’t mean the sponsor segments, but the rest of the video. LTT, MKBHD and all the other tech channels, every movie and game reviewer and a lot of other stuff is all ads. Every single channel that is focused on showing you a new product is effectively an ad. And people watch it because they are interesting in seeing what new or interesting products are out there. There is no aversion to ads, there is an aversion to bad and annoying ads.
- Comment on YouTube is slowing down for users with ad blockers in new wave 10 months ago:
Yes, but that’s self inflicted. They make the rules. They build the software. They decide how ads are presented.
- Comment on YouTube is slowing down for users with ad blockers in new wave 10 months ago:
Because that has been tried so many times over the decades.
When? “Will it Blend?” is about the only time I can think of when a company went in an alternative direction and turned their ads into entertainment and was quite successful at that. How many products do even have as little as an official unboxing video? Stuff like the SteamDeck teardown is what I would love every company doing for all their product. But it’s super rare. Why limit your ads to 30sec fake nonsense when you could have 15min of talking about your product in detail?
They just block everything for the exact same reasons
There wouldn’t be a need or even the ability to block anything if it wouldn’t be forced on the user. If Youtube had a “show me a random ad” button, I’d click it. I don’t hate ads. I hate bad ads that are forced in my face when I don’t need them. I have plenty of downtime where I wouldn’t mind seeing what new products are around. Gameify that stuff. Make it interesting. Make it explorable. Make it interactive. You have million dollar budget, mountains of collected data and random garbage forced into the users face is the best you can come up with?
“What can we do to actually get ANY ad revenue out of this so that we can keep the lights on?”
You are forgetting that there is an advertiser in all this. People that care about getting clicks on ads will have no problem tricking users into accidentally clicking on ads. But why are the advertisers themselves ok with that? If I want to advertise a product I’d not be interested in paying for accidental clicks users were tricked into, I’d be interested in finding users that are interested in the product I want to sell. And I really don’t see current ads doing that very well. They might be better than literally nothing, but I really don’t see them being better than all the potential ways to make better ads.
- Comment on YouTube is slowing down for users with ad blockers in new wave 10 months ago:
I’ll never understand why they spend so much effort pushing ads into people’s faces that don’t want see them and so little making ads more attractive.
A very large chunk of what people consume these days is effectively already ads. Every Youtuber holding a product into the camera is an ad. And people want to watch that. They want to know what new products are out there. It just has to presented appropriately.
Forced ads with mandatory 5sec isn’t making people interested in your product, heck, numerous times I might have been interested in a product, but lost interested since I couldn’t rewind the ad or because the ad didn’t link to anything that gave me further information. A 15min video from a Youtuber reviewing a product in detail is way more effective than any regular ad I have ever seen, yet there are almost no ads in that style.
- Comment on WB’s ‘Ready Player One’ Blockchain, VR, AR, AI ‘Readyverse’ Will Of Course Be A Disaster 10 months ago:
It was never alive to begin with. The Metaverse exists as nothing more than a buzzword made popular by Zuckerberg/Meta. There are no technologies, standards, devices or software driving it. Nobody even agrees on what features it should have.
And on top of that, old pre-buzzword stuff like SecondLife or PlaystationHome are still closer to a “Metaverse” than any of the more recent buzzword filled attempts.
- Comment on Kagi is now partnering with Brave 10 months ago:
Have you ever actually tried it? I only did the trial run, but from my experience it pretty much delivers. Results are pretty much at the level of Google with a lot of junk removed. Nothing else I tried came close. Neither Bing, Yandex nor Brave (all other alternatives are based on Bing).
- Comment on Kagi is now partnering with Brave 10 months ago:
Brave ain’t exactly perfect, but at least they aren’t sucking at the tit of Google Search monopoly.
- Comment on Japan determines copyright doesn't apply to LLM/ML training data 10 months ago:
You can. Distributing copies is illegal, not downloading them. That’s why torrents are bad and streaming sites are fine. (Some exceptions might apply depending on your country).
- Comment on Sony Steals Customers' Purchased Content - Piracy is COMPLETELY JUSTIFIED! 11 months ago:
Is there any platform or medium where I can buy locally stored and DRM-free software?
Steam, but you’ll have to manually search around the forums to see which games does it and which doesn’t. It’s not exactly a well advertised feature, but integration of Steamworks copy protection is optional. But most of the games that are DRM-free on GOG are DRM-free on Steam too.
- Comment on Sony Steals Customers' Purchased Content - Piracy is COMPLETELY JUSTIFIED! 11 months ago:
If the content is not stored locally and DRM free, then you don’t own it.
Have fun managing tens of TB of backups. I have given up on that quite a while ago, DRM-free is just not a practical for the amount of digital content you collect over the years. It’s a nice to have thing that comes in really handy sometimes (e.g. watching movies on unsupported device like VR headsets), but it’s not a solution for digital ownership. In some ways it’s actually worse, as you can’t practically sell DRM-free copies, as you don’t have a proof of ownership.
This needs a legislative solution or some NFT-like thing that gives you a certificate like “You own this, feel free to pirate if we go out of business”.
- Comment on Whats your favorite Main Menu music? 11 months ago:
- Comment on Stable Diffusion XL Turbo can generate AI images as fast as you can type 11 months ago:
Bing Image Creator if you just want to create some images quick (free, Microsoft account required). It’s using DALLE3 behind the scenes, so it’s pretty much state-of-the-art, but rather limited in terms of features otherwise and rather heavy on the censorship.
If you wanna generate something local on your PC with more flexibility, Automatic1111 along with one of the models from CivitAI, needs a reasonably modern graphics card and enough VRAM (8GB+) to be enjoyable and installation can be a bit fiddly (check Youtube & Co. for tutorials). But once past that you can create some pretty wild stuff.
- Comment on Uber slashes fees in Bangladesh as drivers keep taking rides offline 11 months ago:
Problem here is that “Open source platform” is a meaningless term. Open Source is a type of license that regulates how to redistribute source code. None of those principles apply to services and platforms, which are about data and control. If the Uber app would be Open Source, but still had to connect to the Uber server and play by its rules, nothing would change. Meanwhile if it played by its own rules, what rules would that be? There are no established rules for an “open” service platform, especially not when it comes to platform that have review and reputation systems build in.
Simply put, it’s really not clear how you can be “open” yet at the same time provide any level of protection against fraud and abuse.
- Comment on AI Art Generators Can Be Fooled Into Making NSFW Images 11 months ago:
Aren’t there NSFW filters after the generation? Bing Image Creator for example will frequently generate images with a borderline-NSFW[1] prompts, but only show you a subset of the four it generated, not all. Some prompts will also be rejected before any generation takes place at all. But I don’t see how this would help you getting through the filter that happens after the generation.
[1] “borderline-NSFW” really just means anything involving woman or violence, the filter on that thing can be extremely prude.
- Comment on A Spanish agency became so sick of models and influencers that they created their own with AI — and she’s raking in up to $11,000 a month 11 months ago:
Warning! You will never be in a relationship with this model.
It’s AI generated, there is a good chance of them selling relationship-as-a-service in the near future. In a way, your chances with an AI model are way bigger than with a real one.
- Comment on A Spanish agency became so sick of models and influencers that they created their own with AI — and she’s raking in up to $11,000 a month 11 months ago:
How do they ensure the same looking person is generated each time?
You can generate consistent faces simply by using random non-existent names. Wich in turn you can use to train a custom LORA with Dreambooth (needs about 20 images) or use ROOP (a single one can be enough). And of course you can just mix and match it as you please, mix multiple real faces together into a new one, use dedicated face generators like thispersondoesnotexist.com and so on.
This barely even takes effort anymore, e.g. quick ROOP FaceSwap with the photo from the article.
- Comment on A Spanish agency became so sick of models and influencers that they created their own with AI — and she’s raking in up to $11,000 a month 11 months ago:
Horses were never “employed”.
They did a job and got paid for it (in food and housing). Sounds like employment to me. You can call it slavery if you prefer. But that doesn’t change the fact that there were jobs that used to be done by a horse, that is done by a machine now. Meanwhile the resulting increase in productivity and market growth didn’t create new jobs for horses, they were simply no longer needed in the job market. Machines where used for all the new jobs that appeared right from the start. The horsepower the horse provided could be provided easier and cheaper by a machine.
What jobs are left for the human once AI can replace their brainpower? Blue collar jobs might be safe for a while longer until robotics catches up. But a job that is mostly shifting information around, be it spreadsheets, phone calls or art, all of that is slowly getting into reach of AI.
- Comment on A Spanish agency became so sick of models and influencers that they created their own with AI — and she’s raking in up to $11,000 a month 11 months ago:
Body images that are so unobtainable that we literally made them up.
Don’t worry, we already have plenty of beauty filters that can run in real time and make everybody pretty on the Internet.
Seriously, I think people still vastly underestimate how much of everything you see today is already fake. “AI is bad”-news kind of hides the fact that none of this was real to begin with.
- Comment on A Spanish agency became so sick of models and influencers that they created their own with AI — and she’s raking in up to $11,000 a month 11 months ago:
Hasn’t everything been fictional since the invention of Photoshop some decades ago? And even before that people have been faking photos by analog means. Photos, especially in the context of advertisement, have never been non-fictional.