Huge AI Announcement | Stable Diffusion 3 Is A Midjourney Killer Uses Same Technology Found In SORA!

hey everyone how you doing stability AI that's  the company behind stable diffusion just announced   the soon to be release of stable diffusion 3 and  unlike the couple previous versions to me this   one here is going to be a big deal the reason  is stable diffusion 3 is going to incorporate   diffusion Transformer architecture which is the  same kind of Technology that's being used in OpenAIs   Sora so most of us by now have probably seen  videos of open AI Sora so we're familiar with the   capabilities that that kind of technology has and  to bring that into stable diffusion that's going   to be a game changer that's a big deal so this is  big news for all users of stable diffusion some of   the major features that stable diffusion 3 brings  to the creative table are the ability to generate   3D images and video along with greatly improved  text generation and text styling capabilities   so each of these are Big features in and of  themselves but when combined they make stable   diffusion 3 a musthave release so it's stable  diffusion 3 is going to increase the versatility   and functionality of stable diffusion and it's  going to allow stable diffusion to possibly   outperform Dahl-e and even midJourney and put it  on an even footing with Sora since it's release   just a couple weeks ago open AI Sora has been  just killing it so stability AI knew that they   needed to act and they needed to act fast and act  fast they did announcing the release of stable   diffusion 3 just a couple weeks after their  previous release of stable Cascade and stable   Cascade is something I had initially thought of  doing a video on how to install but once I got to   working with it I decided not to because stable  Cascade needs a lot of resources that a lot of   people aren't going to have on their home computer  it needs a pretty beefy GPU something along the   lines of a 3090 or 4090 which you know those are  $1,000 and $2,000 gpus a lot of people don't   have one of those I don't have one of those and I  really only want to make videos first of all that   use technology that's free that's open source  and that really the majority of people can make   use of it's not going to do me any good to make  a video on how to install something if you can't   install it so I want it to be something that's  usable by the majority of people on a standard   computer so I decided not to do a video on  stable Cascade now supposedly stable diffusion   3 Works around that that need for a high-end GPU  by having different variations of the software   some of which are intended to run on Lower  spec gpus now since it's not been released   yet I can't test that but hopefully that stays and  is the case upon release and we will know that   soon enough now some of the major differences and  improvements that you'll find in stable diffusion   3 include significantly better text rendering  capabilities text styling capabilities    the ability to generate 3D images the ability  to do video within stable diffusion and the   ability for the AI to differentiate between  left and right and foreground and background   and that's something that most AI has a hard time  doing right now in other words if you tell your   image creator that you want somebody holding  something in their right hand most AI is not   capable of doing that at this point so that is  something that supposedly is going to be available   on stable diffusion 3 and I'm really anxious to  see if that holds true because that's going to   be very very useful if it does there is a pretty  big concern among the community though that stable   diffusion user community that stable diffusion  3 may be headed towards the direction of being   more heavily censored than previous versions of  stable diffusion have been a lot of people love   stable diffusion for the reason that it is not  heavily censored unlike a lot of the other AI   generators and it gives you creative freedom  to create the art that you want to be able to   create if they start censoring that it's going  to take that ability away and a lot of people are   also concerned about political censorship and  something I haven't seen anybody else talking   about at this point is at least in my opinion I  believe that stable diffusion will probably at   some point become a paid product it just I don't  know that for sure but to me it just makes sense   and I suspect that at some point that will happen  a lot of people would say that it can't become a   paid product because it's released as open source  but they're confusing free with open source which   are two very different things open source simply  means that the source code is made available so   if you want to inspect the source code for  whatever reason for you know if you want to   see if there's anything nefarious hidden within  the source code you can do that but there are   plenty of open-source products that are still  also paid products so open source and free are   not necessarily the same thing sometimes they  are sometimes they're not but there is a lot   of time energy and especially money that goes  into creating something like stable diffusion   and at some point that huge investment needs to  start making a return and I suspect that at some   point stable diffusion will go the route of most  of the other AI generators that are out there and   it will become paid what that should mean to users  of stable diffusion right now is because we have   the ability to use it for free this is your grand  opportunity to get in there use it all that you   can learn everything that you possibly can right  now while it's free before it becomes paid because   then once it becomes paid you already know how to  do everything so you're not trying to learn while   you're also paying for the product you already  know how to use it at that point so what have we   learned about stable diffusion 3 while stable  diffusion 3 is going to have vastly improved   text and text styling capabilities the ability to  differentiate between left and right the ability   to tell between foreground and background it's  going to have have multiple subject capabilities   along with improved graphics and photo realism  so there's actually quite a bit there's there's   a lot to come in the stable diffusion 3 release  the cons are the potential for the high Hardware   specifications which hopefully they will  have a workaround for for people that uh like   me that just have a normal typical GPU in their  computer there's the prospect that it might be   highly censored we won't know until it's actually  released and in my opinion the fact that at some   point it's probably not going to remain free  although I've seen nothing saying that stable   diffusion 3 won't be free so I would assume  that that's a little bit further down the road   if it ever happens at all but those are the  worst case scenario wha ifs so as it stands   right now all of the pros and there's a lot  of Pros greatly outweigh those the cons and stable diffusion three looks like it's going to  be a major release while a major Improvement on   stable diffusion and I'm really really anxious  for it to get here and that's it until stable   diffusion 3 comes out once it's available I  will make a quick tutorial on how to install   stable diffusion 3 and if you found anything in  this video helpful I sure would appreciate the like

Channel: Ron Pierson

Views: 1,418

Keywords: sora, stable diffusion, stable difusion, openai, open ai, ai video, ai text, stable diffusion 3, stable difusion 3, stable cascade, stable diffusion cascade, stable difusion cascade

Id: MLjCDSqT73I

Length: 8min 37sec (517 seconds)

Published: Wed Feb 28 2024