Doesn’t help having contradictory instructions. This will just confuse the LLM and spill out one or the other at times. Though, I would think the model it is trained on would already have an inherent bias against covid disinformation so you’d have to sort of “jailbreak” it into saying something else - which again, doesn’t work like this.
I absolutely hate my upstairs neighbors. Even worse when they have a party up until 4 am. The other day though my smoke alarm threw me out of the bed by violently blaring three times. Never had a false alarm before but always thought it would be if I maybe burn something in the kitchen by accident, not when there’s absolutely nothing going on while I sleep.
An Android rom that emulate the iPhone? It would emulate Android and in either constellation your comment does not make any sense. No one assumed Apple to emulate Android OS.
This could still backfire and then you go from some short additional growth to a potential fall when people actually move away entirely from your products.
But “AI” generated images don’t suffer from too much variety, they suffer from looking samey. It’s the opposite of what you’re arguing about. Limbs aren’t really the issue here since this is about Midjourney, which handles that part fairly well already.
How do you get a quality increase if you by definition cut down on the variety of the generative aspects? That doesn’t make any sense.
Isn’t this causing a huge degradation in quality? It’s like compressing an image over and over again. Those “AI” models can only generate things on what they know, and already have a very real issue of looking samey because of it. So if we train models on that, and then another model on the new model, and repeat this over and over again, we’d end up with less and less quality & variety for each model, no?
In my experience 30 bucks is a terrible price range for in-ears or headphones, mostly Chinese garbage that fails after a few months, maybe a year or two tops, even with just light usage. My current ones were at around 100 bucks and already outlived my previous cheap ones by quite a bit. That being said, FairPhone is kinda well known for being overpriced in regards to cost vs quality too, at least with their phones.
It’s usually quite a hassle to set those tools up, especially if you don’t have much technical knowledge. A lot of the more resource heavy tasks are also not really possible on a home computer and require big servers with multiple GPUs and absurd amounts of VRAM, or very specific APUs but those are still very early. The majority of what you can do at home is typically limited to generating pictures, and even there it takes quite a bit of time if you want some really high quality stuff. For a lot of more complex tasks you’re simply resource limited. And in regards to time I’m just talking for the actual generation process. Getting good results, and to the point of getting them, is another lengthy process that many people underestimate. It’s not a magic button because those LLMs are pretty damn stupid actually.
My issue with the topic is that everyone targets the wrong thing and just jumps on the media hysteria. They are not going to be able to stop the production or distribution of deepfakes, and imo they shouldn’t, because they’re basically just an advanced form of photo & video editing that already existed for decades and it did not bother anyone up until now that “AI” became a media scapegoat. What they should be bother to enforce is the illegitimate use of such material, for things like blackmail, bullying, disinformation etc. Some neckbeards wanking one out on a clearly marked deepfake porn video isn’t really going to harm the person depicted. Using such a video to claim it is real on the other hand to smear or blackmail them on the other hand is. And this type of bullying has also been going on for decades through classical photo & video manipulation, and again, it did not bother anyone up until now. And by focusing on this idiotic media hype instead of the real issue we basically make sure that it keeps on happening, just like with climate change.
Yes. However, the longer we wait, the worse it will get. If we look at all the various emission targets and dates, let’s say carbon neutrality for 2050 or whatever… Then it makes a huge huge different if we now cut 80% of those emissions in the first half, versus cutting them in the later half, or stretching it out painfully until the very end.
That’s why I’m so pessimistic about the whole topic. Everyone just slouches along as if there’s nothing to worry about. Meanwhile, a huge portion of our climate models seem to be even a bit too optimistic and not matching what is happening in the real world. And apparently it still does not care anyone. I just don’t get it.
We’re literally already doing that with greenhouse gasses on a gigantic global scale since the industrial revolution. Anything we can try is better than the course we’re on, especially since absolutely no one seem to want to seriously cut their emissions to 0.
No I get what you are saying, the thing is that the cosmic ray bit flip fits the razor.
It does not, because there’s plenty of simpler and much more likely explanations than that.
I think they even replicated what happened in the video in an emulator with a single bit flip
You think. But they didn’t. They replicated something that fairly looked close, but not the same. If it was a bit flip, it would have been reproducible in the exact same way. You can see this very example being done in the other video someone linked here, disproving your entire argument.
To be honest I kinda suspect you’ve done no effort to fact check this but are just going with your gut feeling?
And I suspect you’re projecting. You watched a video that was spreading the myth and are now hell bent on believing it to be the cause, because you want to, rejecting any other explanation.
I don’t mind discussing this further, but if so I’d really like to hear what your point is, because if it is that: a) cosmic bit flips doesn’t happen or b) a bit flip couldn’t have impacted the game like that then I think you’re better off watching that video I linked or actually read up on the subject because my impression is that if you apply occam’s razor to that mario64 incident… a bit flip is all that’s on the table.
You clearly still miss the point, making me personally not wanting to discuss this further. I know that cosmic bit flips happen, but I also know that the example you’re talking about isn’t caused by a cosmic ray bit flip. It is way more likely that it is caused by some faulty hardware issue than a cosmic ray hitting it exactly at that moment causing exactly this outcome.
Same thing happened during a Mario64 speed run years ago. The particle just happened to be flying through space and the earth and just happened to hit a spot on one of the computer chips. It made Mario warp upwards slipping a crucial section of the clockwork level. They couldn’t figure it out for years.
You cannot seriously believe this bullshit. There are so many other potential possibilities with actually a fair likelihood, than such a stupidly rare cosmic event to be the cause of it.
Another completely useless search engine for me. It defaults to IP based country search regions (like the IP of my default VPN server) with no ability for international searches, and only a drop down menu to switch to other countries instead of a toggle, like DDG.
I don’t think I’ve seen any Threads content on kbin.social nor fedia.io ever. I feel if it was federating I’d see lots of it, no? Since they’re so much larger?
Germany too. And Threads isn’t even part of the Fediverse yet afaik anyway?
Gab is another far right social media site and I guess they implemented “their own” chatbot, which is definitely not GPT-4. https://en.wikipedia.org/wiki/Gab_(social_network)