It was the Parthenon, not the Pentagon.     See Less

It's called the PARTHENON in Greece and one of the oldest ancient buildings, not the PENTAGON!!!     See Less

That’s Siri for ios18     See Less

The interesting thing is Chat GPT 3 was released in November 2022 while it took 5 months for Chat GPT 4 was released in March 2023. It's been over a year for Chat GPT 5 hasn't been r     See More t but is looking around mid-summer. I wonder the reason why it's taking a bit for Chat GPT 5 is they're really taking their time to fine tune the A.I.    See Less

Sora didn't just change the characters, parts of the background also changed, if you check the dustbins, wall, graffiti, large bin etc.     See Less

User
please write a 10 minute stand up comedy set about current ai advancements
ChatGPT

[On stage, under the spotlight]

Hey, everyone! How's it going? So, I've be     See More g about artificial intelligence lately. You know, AI, the technology that's supposed to make our lives easier, but also occasionally makes us question whether we'll end up being ruled by robots. Yeah, that one.

Have you noticed how AI is everywhere these days? It's like that one friend who suddenly shows up to every party uninvited, but you can't help but enjoy their company anyway. You're at the grocery store, and AI is there, helping you pick out the perfect avocado. You're watching Netflix, and AI is there, suggesting movies you didn't even know existed but end up loving. It's like having a really helpful but slightly nosy roommate.

But let's talk about some of the recent advancements in AI. They're doing some pretty impressive stuff. I mean, they've created AI that can write poetry, compose music, even paint pictures. It's like AI is going through its rebellious teenage phase, trying to prove it's not just good at math and science but also has a creative side.

And then there's AI in medicine. Have you heard about this? They're using AI to diagnose diseases, predict patient outcomes, and even assist in surgeries. It's like having a super-smart doctor who never gets tired and never forgets to wash their hands. Although, knowing my luck, it's probably the same AI that keeps suggesting I eat more kale.

But of course, with all this advancement comes some... let's call them "quirks" of AI. Like those virtual assistants on our phones. You know the ones I'm talking about. You ask them a question, and half the time, they give you the most random answer imaginable. You're like, "Hey Siri, what's the meaning of life?" And Siri's like, "Life is like a box of chocolates... oh wait, wrong movie."

And let's not forget about self-driving cars. Now, I don't know about you, but the idea of putting my life in the hands of a car that's basically just a really fancy Roomba terrifies me. I mean, I've seen how my Roomba navigates around my living room. It's like watching a drunk person trying to find their way home from a party.

But hey, at least AI is helping us tackle some of the big problems, right? Like climate change. They're using AI to optimize energy usage, predict weather patterns, and even develop new materials that are more eco-friendly. It's like having a whole team of environmental superheroes, except instead of capes, they're wearing pocket protectors.

But you know what's really impressive? AI in language translation. Have you ever used one of those apps that translates text in real-time? It's like having your own personal Babel fish from "The Hitchhiker's Guide to the Galaxy." Now, if only they could make one that translates my awkward attempts at flirting into smooth pickup lines. "Hey baby, are you a software update? Because you've got my heart racing and my palms sweaty."

And let's not forget about AI in entertainment. They're using AI to create virtual actors, like those deepfake videos that can make anyone say anything. It's like living in a world where Forrest Gump can give you financial advice and Darth Vader can teach you how to cook lasagna. Although, let's be honest, we'd all watch that cooking show.

But with all these amazing advancements, there are also some... shall we say, ethical dilemmas to consider. Like, what happens when AI becomes so advanced that it starts developing its own sense of consciousness? Will we have to start treating our smartphones like they're our new overlords? "Sorry, Siri, I didn't mean to offend you. Please don't shut down my bank account."

In conclusion, folks, the future of AI is both exhilarating and slightly terrifying. But hey, if it means I never have to fold laundry again or figure out what to cook for dinner, I'm all for it. Just as long as they don't teach my Roomba any new tricks. Last thing I need is for it to start demanding a raise and organizing a union.

Thanks, everyone! You've been a great audience. And remember, if a robot ever asks you if you're a robot, just say no. We don't need any more existential crises in this world. Have a great night!    See Less

26:00 one guy crack the model, and every photo cant be trusted anymore     See Less

Looks like a ton of the details were changed in the backgrounds of those character replacement videos; so I suspect either the whole clip was generated from prompt only (and just changing th     See More r description), or there was a original video input but the system didn't just mask the region where the character would be but regenerated everything.    See Less

Are you sure that you can easily take your own footage to modify it with that result?

I would say that this "simply" works like StoryDiffusion which works like Stable Diffu     See More In Stable Diffusion you can create that effect if you first create the image you like with Text2Image.
* Then you keep the model, prompt, parameters, Lora's and especially important the seed and take them to Image2Image.
* It's very important that you have described your character and also the background as detailed as possible when you created it in Text2Image !
* In Image2Image now you keep everything the same but only change the description of the character AND set the CFG scale to a relatively high value like 10 - 15 to follow along with your prompt and set the denoising strength to something between 3.5 and 5.5 - This gives the model some freedom to not stick to close to the original image, but enough to change the image related to the prompt! - So CFG scale and Denoising strength go hand in hand and it takes a little of trial and error to achieve the result you like.
Because you only changed the prompt for the character the background will vary a little but the character will adapt more to your new character prompt.
You have to play with the ratio between CFG scale and Denoising strength!

Sora is not a model but an application so there's nothing that restrict's the app programmers or even users (if made possible for them) to not use inpainting on top of that to amplify the character even more!

It's just a guess, but I would program it that way: If you repeat that for every nth keyframe and then recalculate the keyframes between, then you should get that result: A new character in a slightly modified background. - And with that workflow it must be that result because if you would keep the Denoising strength lower, the background would still be very consistent, but the character would have no freedom to develop into the new one.
If you wanted a consistent background, you would have definitely to first keep Denoising strength lower and then do a few passes of inpainting with increased Denoising strength between 6 and 8 to form your character. You then would first go high with the Denoising strength and with further iterations you would lower it to get finer details.
That takes a lot of compute and time for every frame but givers great results.

I do not know if it works that way but if I look at the results, it seems to me that that's the basic workflow of that feature.

You may try that workflow and play with it on your own with StableDiffusion.

If you now could take your own footage directly into Image2Image you would have to describe the image very precisely and would have to know how your model would adapt your prompt to get a pretty close image on the first run so you can preserve the seed for the complete workflow. - In my experience that's very hard to accomplish!

But all in all very impressive!    See Less

If I were a Hollywood actor , Id'e be VERY nervous ......Also for the general public , i fear the deep fakes being used against us by criminals . Is there a way that ALL AI visual mat     See More be encoded with a watermark of some kind and a beep with audio that can NOT be removed by creators ?    See Less