Control net + better prompt coherence + improved char and style ref quality/resemblance. That's all we want. The aesthetic quality is so good already. They really need to work out on these details that will allow people to create more complex projects.
thanks for the summary... I think Midjourney should focus on Images not on Video. There is regarding FLUX already some big Competition happening, and most of all > Midjourney should get their Editor tool running properly, there is too much moderation bugs and also sometimes i don´t even get anything added to a image when erasing and prompting someting into..anyway keep the good work up.
I've screamed it before, and I'll keep screaming about it. They _need_ to hold themselves _accountable_ and help themselves with a roadmap by sticking to some type of time-based release model, or at least rolling out smaller upgrades frequently. The clients (us) _and_ them have to feel TRACTION. Right now, there isn't much. I'm sticking to them more out of loyalty than their product right now. Thanks as always for helping us out with these updates, man. 💯
Interesting comments about price versus performance. I guess the current quality of AI tools isn’t simply a limitation of the technology but rather a limitation of how much customers would be willing to pay. Can’t make it better if no one could afford it.
@ I can’t help but wonder if the AI bubble is destined to burst. Seems like most AI companies survive on funding rather than profits, because it’s just too expensive to operate. And if the only way to advance AI is by adding more and more gpu’s, are we going to cover the planet in AI datacenters? And then what, build them on space stations or on moon? And then there’s the enormous power needs of AI. Tech companies are literally buying nuclear power plants to power AI. It all just seems nuts to me.
Thanks for the update!! I am more interested in World Building than video right now (because video generation is just too short and wonky), but I don't understand how the storytelling fits in. How will that work? Other 3D world builders like World Labs demo from just yesterday take a single image and create a world you can walk in (for only a few feet right now, but hopefully more soon).
Wow, mj video sounds awesome... Been waiting 2 years for control net and outline imputs. So so many companies are using it. Strange that was never top of the list. . Im working on a man hiding a bottle of wine behind his back. Just so so difficult to do :(
It's a letdown to hear it will be a simple video model. There are already so many of those. But if they can kill all glitches and artefacts from the existing models, they would be on a good way.
Control net + better prompt coherence + improved char and style ref quality/resemblance.
That's all we want. The aesthetic quality is so good already. They really need to work out on these details that will allow people to create more complex projects.
Well said!
thanks for the summary... I think Midjourney should focus on Images not on Video. There is regarding FLUX already some big Competition happening, and most of all > Midjourney should get their Editor tool running properly, there is too much moderation bugs and also sometimes i don´t even get anything added to a image when erasing and prompting someting into..anyway keep the good work up.
And thank you for sharing your thoughts!!
I've screamed it before, and I'll keep screaming about it. They _need_ to hold themselves _accountable_ and help themselves with a roadmap by sticking to some type of time-based release model, or at least rolling out smaller upgrades frequently. The clients (us) _and_ them have to feel TRACTION. Right now, there isn't much. I'm sticking to them more out of loyalty than their product right now. Thanks as always for helping us out with these updates, man. 💯
And thanks for sharing your thoughts!
Thanks! Missed the Discord this week, so..... appreciate!
haha I'm glad I could help!
When is Editor open for all subscriptions?
Honestly no idea!
Interesting comments about price versus performance. I guess the current quality of AI tools isn’t simply a limitation of the technology but rather a limitation of how much customers would be willing to pay. Can’t make it better if no one could afford it.
Yah it would be wild to see what could be unleashed if you went full Jurassic Park on it and spare no expense
@ I can’t help but wonder if the AI bubble is destined to burst. Seems like most AI companies survive on funding rather than profits, because it’s just too expensive to operate. And if the only way to advance AI is by adding more and more gpu’s, are we going to cover the planet in AI datacenters? And then what, build them on space stations or on moon? And then there’s the enormous power needs of AI. Tech companies are literally buying nuclear power plants to power AI. It all just seems nuts to me.
We can always count on you
Just trying my best!
Would RNDR help speed up rendering? Might be a good partnership.
That's above my paygrade 😂
I hope they focus on Images first. The moodboard, making one owns style sounds really good. What does worldbuilding mean?
Nobody knows!
Wonder when I will get the canvas. I'm a paying customer but i still don't have it. Bad policy.
That's my question too, still hoping I get to use the feature soon
I really don't know but I can say I hope it's soon!
Thanks for the update!!
I am more interested in World Building than video right now (because video generation is just too short and wonky), but I don't understand how the storytelling fits in. How will that work? Other 3D world builders like World Labs demo from just yesterday take a single image and create a world you can walk in (for only a few feet right now, but hopefully more soon).
They need to sort the weird fisheye effect, looks very early unreal engine
Yah I want to see world simulations too!
Wow, mj video sounds awesome...
Been waiting 2 years for control net and outline imputs. So so many companies are using it.
Strange that was never top of the list. .
Im working on a man hiding a bottle of wine behind his back. Just so so difficult to do :(
oh man that does sound like a challenge!
Hmm so no info about niji 7?, because in mu current workflow i mostly using niji than mj
I'm pretty sure the niji team is working on video and MJ V7 will be able to do everything Niji can ... or maybe niji will be like a 'style' for MJ
@ i see so in the future, probability niji will be absorbed as style option in mj
Thanks for your unrelenting effort, Nolan!
happy to help!
It's a letdown to hear it will be a simple video model. There are already so many of those. But if they can kill all glitches and artefacts from the existing models, they would be on a good way.
I kind of agree!
Just give me flux
👋 hi
🤝