Luma Labs Stunning "DREAM MACHINE" Is Bigger Than You Think!
Вставка
- Опубліковано 16 чер 2024
- Learn A.I With me - www.skool.com/postagiprepardness
🐤 Follow Me on Twitter / theaigrid
🌐 Checkout My website - theaigrid.com/
Links From Todays Video:
LUMA - x.com/CuriousRefuge/status/18...
x.com/LumaLabsAI/status/18009...
x.com/amebagpt/status/1801234...
Welcome to my channel where i bring you the latest breakthroughs in AI. From deep learning to robotics, i cover it all. My videos offer valuable insights and perspectives that will expand your knowledge and understanding of this rapidly evolving field. Be sure to subscribe and stay updated on my latest videos.
Was there anything i missed?
(For Business Enquiries) contact@theaigrid.com
#LLM #Largelanguagemodel #chatgpt
#AI
#ArtificialIntelligence
#MachineLearning
#DeepLearning
#NeuralNetworks
#Robotics
#DataScience - Наука та технологія
I've tried it, its not perfect!!!!..
But i appreciate it for being open, this gives the developers more help in developing the training model properly.
Agreed. It’s good to know this is version 1, but version 1 had only one decent video out of 15 prompts. Their marketing materials may have been created with more compute than we have access to.
@@earth_safariI had several good prompts, I think your luck will vary depending on what your trying to generate
@@earth_safari i found turning off the 'enhance prompt' works a lot better
I know this was uploaded a little late but this software is amazing, there was also an experiment which i didn't get to cover where they had a box with the word "apples" on the outisde, then the prompt was "opening the box" and then inside it showed apples, was interesting to see their world model
Now here comes Runway Gen 3 😂 at this rate, Sora should be able to create a short film upon release. If they’re waiting this long.
I use this tool to animate a picture of my dead mother
it was scary as fuck
i hope tools will get better so I can see my mom again
I'm sorry for your loss.. In our lifetime you might actually see her again in a virtual world :)
My mom passed away in the last year, to your point I kind of want to do this just to see. Not that I think it’s a remedy for grief, but it’s a pathway to explore grief.
Ive already started integrating it into my content. its amazing once you understand the limitations based on what you are trying to do.
Oh, now THIS is something i wanna' play with. Thank you!!
I Hope The Free Use Feature Stays. It'll Make It Way More Accessible In My Opinion.
Along With That I Hope The Developers Will Manage To Extend The Video Lenght From 5 Seconds To 10 , 15 , 20 And So On As The Site Evolves.
Im watching more of AI Search, as they use timestamps...
I tried it, it was impressively good compared to any so far. Their 3D modeling is also amazing
"The Level 2 of AI Adoption is arriving before AGI"
The Last AI of Humanity (New Book)
12:30, Luma's quality does go up when you use image to Video, But that depends on the quality of starting video and how coherent Luma can stay to the video by the end. I've seen some amazing flawless anime clips come of Luma but many not so much. But we have yet to see Sora do anime and that's a shame.
It did make my prompt vid, something like a day later. Same sort of morphing issues Stable Diffusion has.
Best review of luma so far
You make really amazing informative videos. I have also deep passion in ai to education and inform people about this rapid changing world
Any way of you expanding your free membership for your new program you started, missed out on the first 200 , I may message you this a few times, thanks in advance
Are they saying how big the model is?
Hooray !
non soap opera content !
😅😂
8:20, The usable videos might go up when the sites traffic goes down and able to decorate more compute to their videos, But the way I'm looking at the paid tier list is the Standard gives you 150 generations a month I'm assuming that's all at once. I can see my self blowing through those in a day if I worked at it. Then what is the service doing for me for the next 28-30 days until the next month. I'd like to see them implement daily generation credits like other sites, like Standard could give out 5 credits a day the could accumulate over time and cap out at 150. That still comes out to about 150 a month.
Yep, for the amount of errors it outputs without the ability to preview anything, the least they could do is add in some free dailies or a 'relaxed' mode for paid tiers. Their UI needs a lot of work as well.
It is time to research lucid dream therapy & your humanity is off score.
Lucid dream therapy can potentially be used to study and induce out-of-body experiences(OBEs) and autoscopic phenomena (AP) in a controlled setting.
EEG and neuroimaging studies on lucid dreamers could elucidate the neural underpinning of OBEs/AP, which involves regions like the temporo-parietal junction.
Overall, the ability to volitionally recreate and study OBEs/AP within the controlled context of lucid dreams opens up new avenues for research and therapeutic interventions.
A white tiger jumping through a ring of fire at the circus: ua-cam.com/video/yVAK_rrS6Sc/v-deo.htmlsi=pE88wElL7fnddP1e
Well all the dreams i create just fails.... so nothing really works very well currently....
One like in dream
Not even close to Sora
well Sora effectively doesnt exist until its released.
Forget sora ..they also dont even close to google vidu.
@@piteshbhanushali1140 another ghost product.
I wont disagree on quality, but one has to be pragmatic and realize those are not part of the conversation because they are still in a lab somewhere
I thought Sora and VIDU were fake. Especially after that balloon video and finding out that they actually didn’t generate the whole video and they actually used the person.. I think it’s all fake marketing hype to get their stock up
@@kevinsm2039 Not fake it's throuly AI, look at the videos
Tried it for the last few days. It's awful. Great for being something that's open and allowing us to try but the quality is really bad. Definitely plus points for being accessible though...
too late, another SOTA model has just been dropped 30 minutes ago!
The image aka ai generated are akin to little girls ken and barbie figurine . Eyes do not change car do not move , the image move . It's a nice trick , it's how creation around earth do it .take fastest velocity of earth solar system milky way this galaxy . If something is free in atmosphere . The free will keep going (resistance to change)everything work like this , yep even light . Light also do not move . Everything around move . Combined with light cloning on various medium once generated in a spherical way and you are in business . No medium in space right ? So human nose can smell nothing then , how so ? If there was no aka filter on earth ? The sun ouput smell of burning even tho space is claimed to be nothing . Nothing does not smell like burning . So there is a medium . And its a very specific path a small drillbit wobble path
Bigger than you think? It literally feels like where I thought AI was 3 years ago... its terrible lol
It’s the worst it’s ever going to be
it is neither consistent or accurate
Its only good for image to video..text to video sucks..