LLM is AI. So are NPCs in video games that just use if-else statements.
Don't confuse AI in real-life with AI in fiction (like movies).
LLM is AI. So are NPCs in video games that just use if-else statements.
Don't confuse AI in real-life with AI in fiction (like movies).
Despite how bad Google Search had become, DuckDuckGo and Bing are somehow still worse. While Google displays the result in the first few, DDG and Bing have no idea what I'm looking for.
Gotta try Kagi sometime.
Based on my understanding of how current diffusion models work, you actually don't need to train it on CP. As long as it knows how humans look like without clothes and how children look like even if fully clothed with abayas and stuff, it can make the relation and generate CP when asked to.
Just to be clear, I'm totally against any form of CP and CSAM. Just explaining how the tech works.
I always do. I'm a fan of JetBrains Mono.
WebP is awesome (JPEG-XL is awesomer though). It compresses better than JPEG which was introduced 30(!) years ago. It's time for JPEG to go away.
This is what AI actually is. Not the super-intelligent "AI" that you see in movies, those are fiction.
The NPC you see in video games with a few branches of if-else statements? Yeah that's AI too.
IMO the focus should be on lowering the prices. A lot of people in my country still rely on spotty mobile data as their primary internet. Imagine 100 mbps fiber for $10 a month, that would be awesome.
This smells like investor-baiting. Studios don't really need to announce that they're going "aggressive" in using a certain tool.
25/3 is perfectly usable for a single user, provided you don't need to upload stuff. Watching 1080p60 on YouTube only needs slightly over 12 mbps.
I'm not defending the current state of the internet services, just saying it's not that bad.
If you make it reproduce copyrighted media, it is a problem.
As long as the stuff it generates doesn't resemble any copyrighted works, even if it was trained on copyrighted works, I don't see why that should be problem.
Exactly
And worry about other shit that you have to deal with
People will say the amount they get paid is miniscule but I'd say it's better than nothing when you use an adblocker. Plus I also get YT Music.
If you do this and then later you face some sort of issue with Windows, remember that it might not be Windows' fault.
Only because I get both YT Premium and YT Music for cheap ($4/month).
They hardcoded it because there were no mod tools back then, and the repo was basically lemmy.ml.
I don't think we should look at this anymore.
Malaysia is .my
I love Linux and Windows 11 (I use both daily). It's quite tiring to see that the usual discourse about Windows is to switch to Linux, instead of actually about using Windows.
Uhh no? They've gotten really cheap nowadays.
$300 gets you a 5G capable phone with a high refresh OLED and a processor fast enough to play resource intensive games like Genshin Impact.
How do I install my Game Pass games?
This is why I run local uncensored LLMs. There's nothing it won't answer.
I can agree but with two conditions. Benchmarks must always be done in native resolution. Hardware capability / system requirement must not take any upscaling into account.
For example, if a studio publishes the requirements for playing at 1080p, 60 FPS, High RT, it must be native 1080p and not 1080p with upscaling.
Disagree. It's only unethical if you use it to generate the artist's existing pieces and claim it as yours.
Of course I do. It's the most convenient desktop OS I've used.
Windows 11 is just Windows 10 23H2. It's just a number. Nothing stops MS from dropping support for older processor in an update for Windows 10.
I'm still using both. Lemmy is my main but for communities that are non-existent on Lemmy, I revert to Reddit.
Disagree with milk. It can stay in B because it makes cappuccino and latte possible.
If you disagree, it means your opinions differ than mine.
That's way too expensive. I guess regional pricing did me good because here it costs an equivalent of $3.75 (RM 17.90). That much for Premium + Music ain't no way I'm not taking it.
Spotify costs $3.35 (RM 15.90).
Training on copyrighted data should be allowed as long as it's something publicly posted.
"leak training data"? What? That's not how LLMs work. I guess a sensational headline attracts more clicks than a factually accurate one.
Not even 5 ms. I have a properly set up Wi-Fi at home and you'll feel no difference in gaming. Wi-Fi only adds like 1-2 ms latency at most.
I love Starfield and has been playing it every day since launch. It runs like dogshit. Sure it doesn't stutter or anything but I can't, for the life of me, get the average FPS in outdoor areas to be anything higher than 70. 5800X + 3080 Ti. It doesn't matter how much I lower the setting, the CPU overhead is crazy.
Let's not bring that X/Twitter shit to Lemmy.
Disillusion?
If it's fully remote, yes.
women
Stable Diffusion is open source. LLaMA is open source.
Support those and not Midjourney/OpenAI/Bard/etc.
If you took a random concept and explained it to a person they could using their existing knowledge set, draw it somewhat competently. That is because people are able to apply knowledge to make something new.
Theoretically it can, but it would involve meticulous and proper labeling of each training data. Currently most of the trained data are automatically labeled and they're not descriptive/verbose enough. I believe the improvements from the latest version of DALL-E is due to OpenAI's use of a more advanced image labeler.
If you knew the exact terms to get the the AI to recreate something in its training data, it could, 1:1.
That's because you told it to. Don't make it recreate existing art then.
And if you ask it to create you something new, no matter what parameters you use it will look like a mess of garbage data.
This is not always true. You can train it on a certain style and a photo of a random object, then have it generate an image of the random object in that style. It will "understand" the concept of a style and an object.
ultimately all the capitalists running these tools see it as is another method to bring the public under their exclusive and totalitarian control.
Exactly why I'm not supporting the closed source paid services (Midjourney, ChatGPT, Bing Chat, DALL-E etc.) and instead advocate for open source projects like Stable Diffusion and LLaMA.
Don't talk to me if your average ratio is less than 1.0.