Out of 8 videos none of mine useable, all had odd morphing. The demos they are most main video creators showcase are taking their handful of best out of 100s maybe 1000s of rolls
Probably hundreds. But if anyone is a seasoned vet of using AI, they should know these are A - Cherry Picked and B - they should expect to run a dozen or more generations to hit a desired result.
@@kaelkent2496 No, it is not unusable. You get a refresh every week. The only limit you actually have to deal with is how frequently you generate. Space it out over the week and use your generations efficiently. It is not hard... and it is not unusable... Maybe just be less picky and put the work in like the rest of us.
@@mythicproductions9154 seems like u focused only on the word unusable which was one part of my comment, i said "UNUSABLE ON THE 15 DOLLAR PLAN", and i stand by that , you need the unlimited plan for it to be usable. before calling me picky how about u read the whole comment. Gee wiz the 15 dollar plan you can easily run through 8 generations because of the small credits u get , so next time before you call someone picky how about you read the whole comment. Anyone considering doing this professionally by the way don't have time to wait a week for regeneration, no thank you, i would just get the 95 dollar plan if i want to WORK as you say. i work in video editing bud, how about you get off your high horse because you sound very condescending. The 15 dollar plan is a guinea pig plan for this company, and a waste of money, u need the 95 dollar plan for this to be worth it period. I've been doing ai several years now, i have made professional projects with it at work and compared to other ai i have used thats more consistent, yea the 15 dollar plan is not worth it.
@@mythicproductions9154 don't know why my comment disappeared but, no its not completely unusable, i said the 15 dollar plan is unusable, from the perspective of someone who works in editing, only the 95 dollar plan would be worth it, no one who works professionally would want to wait a week for regeneration of credits on deadlines, and to say im picky how about you get over your assumptions and condescending attitude when u didn't take the time to read my comment and focused on only the word unusable, i didnt say its completely unusable, i said the 15 dollar plan is unusable also people like u who don't advocate for better than subpar and are happy wasting 15 dollars for a subpar product will allow companies to keep taking advantage of you. People kept braggin on adobe and then now all of a sudden that they see the price model and the terms of service that they can train on your work in adobe now want to fight for rights when it should have been done long ago. I will advocate for improvement that this is not acceptable on a 15 dollar plan.
i agree with you bro, its not worth it on the 15 plan its unusable. the 95 dollar plan is the only plan that's usable for something professional with deadlines. IN REGARDS TO THE user mythicproductions9154 The $15 plan is what I found unusable. As someone who works in editing, only the $95 plan is worth it. Professionals can't wait a week for credit regeneration on tight deadlines. Your assumption about me being picky is misplaced. Instead of focusing on just the word 'unusable,' consider reading my comment thoroughly. Accepting subpar products and wasting $15 allows companies to take advantage of you. Advocacy for better services is crucial. I stand by my point that the $15 plan needs improvement.
Gen-3 is pretty amazing. Content moderation is a bit hyper sensitive. It blocked this prompt a couple of times. I think because I references "exposed" wiring. Here is the prompt. It was part of a sci-fi horror short version 0.1 I posted using gen-3 where I am going to show my progress and improvements on the short. Prompt: " Wide-angle tracking shot: The camera sweeps through a damaged hallway in a futuristic space station. Exposed wiring sparks, with fire and smoke erupting from wall breaches. Automatic doors glitch, unable to fully open or close. The scene is illuminated by erratic emergency lighting." They will get that part right I hope.
I was able to visualize it really easy while reading it. well written for humans. hope gen3 gets it. - if "exposed" is causing you trouble maybe try other wording for just that part
@@animalaliens I shared this and some others that were blocked. They agreed they should not have been and are fine tuning the moderation. Safety first and I agree but it is still in Alpha state and they will get it right at some point.
I made a music video with Gen-3 and it cost me about $30 in credits. As a music producer and artist using it in the real world, I would say it's worth it.
Having used a similar amount of credits, i have to say I question your judgement. It is a novelty atm. If people watch your music video 10 years from now, what do you think they will think of it? How well do you think they will think it has aged?
@@Kaotonix Watched it (assuming it is the one with the waves and the fish). I still don't think it's a worthwhile tool at the moment. Too bad prompt comprehension, too chaotic (not your video, the tool - I can see you did enough editing to temper that). For now, I would rather use something like Luma, or just run stable diffussion locally. I couldn't really recommend runway v3 as it is. Image quality is great though. If they allow img to vid, may be different. Anyway, that video was better than I expected, and music not AI generated (thought it would be someone who "made" a track in suno), so props for that.
@@bjorn0helander thanks for taking the time, yes I’ve been making music for a decade and never really had a great tool for visualizing the sonic soundscapes I spent so much time crafting. I def spent a lot more money on previous projects so I knew what I wanted with this one and had experience prompting, which is how it cost me $30. Honestly I tried luma and others, including gen 2. Def did not yield anything worthwhile. For me, gen 3 was the first time it felt useful. Anyway, cheers 🥂 we have an exciting future for video generation. I tried stable diffusion but my graphics card is shit so didn’t work out for me
The *keyframe* tool in *Luma* would be much more useful if they add these *2 features:* *1.* Ability to add multiple keyframes. I want to add more than 2 images, because an action may have multiple steps to it. *2.* Transition between frames using camera effects. I want to be able to choose a camera effect to apply when transitioning from one frame to another.
I can give you a tip for that. First you use your 1st 2 images. when you want to download your animation, you will see that you can also download just 1 image if you want. So you could download just the last image of your animation to calculate a new animation with that image as Input and you can add an other image as output image. It works very well
Layers and exr files = essential Control = Essential The input should be a 3D app and the final out output should be a composite. The goal should be to create render layers for a final result
I made a music video with Gen 3. While using it this can be very frustrating experience but after some time you learn the limitations of Gen 3 and get some nice results which are better than Luma or Haiper which I used before. You also need the unlimited 95$ plan.
I just made my first AI film using Runway Gen-3 after trying Luma. Wasn't sure I could handle being limited to text prompts after seeing how amazing Luma is at turning images to video. But once I got used to Gen-3 prompting, I actually liked it better. You get to be a movie director rather than just trying to get images to do something interesting (and mostly failing at it).
I have some establishing shots that I used in Gen-3 and was impressed how downtown Tokyo looks at midnight. I wish I could upload pics tho and I would like to pick my aspect ratio.
A lot of this looks to be rendered in slow motion. Does anyone know the default FPS (frames per second) and resolution? I realize that version Alpha has limited control and functionality, but much of the action in the background is really unusable unless the application is comedy, at least in the examples you've shown here.
i believe it is intentional . there should be a problem with integrating physics and and angles in real time . they hide this glitch by going with slow motion.
I'm extremely disappointed with the pricing and quality of Runway Gen-3. The cost is high, especially considering the wasted generations and poor physics, which make it almost worthless for the average consumer. This product shouldn't cost more than Midjourney on a monthly plan given its current issues. The price doesn't reflect the quality at all. Unless you're a business, it's hard to justify the expense. The pricing structure is a major letdown.
Are you forgetting that price is mostly decided by their cost to run the service at this point? You're demanding the impossible.. Reminds me of the "angry man shouting at the sky" meme
@@pictzone More like the old dehydrated dying man in the hot desert sands and the bottle of water is just out of reach. If these tools cost just as much as traditional filmmaking, why would anyone use these inferior tools. AI was supposed to empower the everyman, not just the rich production houses flush with cash. At least Luma gives us 30 free generations per month. Runway needs to throw people a bone.
@@pictzone Dumb. They're complaining about value (which is valid). Value has nothing to do with price. Why pay 10 times more for something that is twice as good.
@@VJ4rawr2 So they should finance small creators out of their pocket right? Have some patience, this tech will be affordable in less than 6 months. Be thankful it's "relatively" afforadble right now, although with limited control on the creative output. The 100$ package is very generous for what it provides in my opinion (check it out if you don't know what I'm reffering to)
It's true - these tools (especially when they come out) have quite a pricey tag. However, with more competition these prices will change (or else a competitior will take the market). These tools are still very new so let's just keep an eye out. We certainly understand the prices hurting the average users at this point!
I must admit that I prefere Luma than Ruway for the moment and it is cheaper, and it gives very good results with Input and Output pictures very quickly.
its TOTALLY WORTH IT on the unlimited plan, on credits alone no it may not be but i would HIGHLY encourage anyone interested to get unlimited for the month the renders are a tad slower but you can play around as much as you want and to me the quality beats anything else right now you can get the prompts down pat but it may take more then one generation to get the output you want so thats why again i say unlimited is CRITICAL for anyone wanting to use this unless you just have a ton of money to throw at it, otherwise you WILL burn through credits FAST for sure!
Did Runway AI's gen 3 updated again? on their pricing modle the lowest paid plan says "Gen-3 Alpha Turbo (Image to Video) up to 10 sec"??? Or i am guessing it doesn't have image to video for you because maybe it was the free version you were using?
I really like your videos. But I am quite lost in it. Pls, what AI platform could work with a real video then parexample to keep a real persons of the video then recreate a background or only create an artistic things abound those real persons? I mean in a very good quality. Thank you very much.
@@curiousrefuge I did. It is a sh.t, a fraud. One generates a video with no sure results but payed ofcourse, it doesn 't follow the prompts at all, even the easy ones. It is a joke.
Given the results achieved, this is my idea, they shouldn't ask for money. When they have reached a decent quality, then it is right that they ask you to pay.
So... who is paying for the servers to run and generate everyone's content? Money does not grow on trees and Runway is not exactly as rich as Elon Musk. Second, if you are familiar with using Ai at all, it goes without saying that you HAVE to cherry pick for the best results. It is a part of the process.
@@mythicproductions9154 I know well that the company that manages the service has to pay for the use of the servers, the programmers...etc. But there is also something called: business risk. An alternative could be crowdfunding, which seems more coherent to me for this type of initiative.
Is there any indication that if you get a result with bizarre artifacts or physical impossibilities, you can re-roll more times, , adjust the prompt, and finally get it to look completely realistic?
Yea... forgive me but that is how prompting works. From the first generation and onward you are either tweaking the prompt or re-rolling results. How much you need to do either varies. I have personally taken on a workflow where I prioritize based on the following: Art Style -> Visual Tone -> Setting Description -> Included elements (People, items, etc). Video it can be a bit harder but I run with the same practice in mind. Another habit I started is before I prompt for a video (if I can't or don't have ref images), I will use the prompt in an image generator to get an idea of the results I could expect, dome some tweaking, then take it to Runway. If I can get what I want well enough in images, I will use a keyframe tool like KREA to generate the video instead.
I tried the tool straight away because the examples look fantastic. The results were more than disappointing. At first I thought it was the prompts. Then I tried a few original prompts from the website. Unfortunately, they rarely produced the same result. They really need to improve that.
We've heard similar stories from other users. It seems it could be do to the extreme use by lots of people at the same time (stress on servers?). Not sure if this is true, but it seems like people are having much better luck now!
Paid subs on Luma take 2-4 minutes and since there is no free version in Runway (yes, the free version of Luma can be quite slow), how about an apples to apples comparison and ditch the bias? That's a pretty big detail to miss so what else is being skewed in your reporting?
Runway does have a non-paid tier. About 130 seconds to be specific. It is not amazing but it is not bad. And it is common practice with most AI tools to reserve their latest releases to their paid clients first before going full problem. Open AI operated like this. Runway Operates like this. Most do. These are well known aspects of the AI space sense last year if not earlier. He did an apples to apples comparison to the capabilities of both products. If you don't like that he didn't review business model details, grow up, stop being lazy, and go check out the platforms yourself and read. This subject was touched on when Runway first released a long time ago and their business model has been the same sense. There is no reason to review it again without a worth while notable update.
@@mythicproductions9154 Not for Gen 3 as far as I can see. (only for gen 2). The issue is: Caleb is implying that while Runway's paid tier is fast, Luma's is slow and that couldn't be further from the truth. Luma's free tier is slow, but Runway Gen3 does not have a free tier to compare that too (I wonder why that is? Would it too be slow?)
@@markrichards5630 Well, yes and as I pointed out it is standard practice for a business like Runway to roll out these updates to paid users before they are open to general free use. Idk how long that will be but that has been the practice in the past and that is what I expect going forward. I can not comment to the gen speed of Luma, but other factors could play into this. High user volume will slow a server down. If he is hitting long delays in generations, it could be because of user volume, internet connection, interfere, and other issues I am not sure I am thinking of. My only problem with the OG comment is the jab and assumption that what was covered in the video is with bad intent or at least that is how it sounds. Also, might I add, it is borderline impossible to expect someone like him in his place to not be biased on a platform. Regardless of what a content creator may say, it should go without saying everyone should do their own research and make the decision for themselves which platform they prefer.
We may be the only channel (or relative to our size) that doesn't take sponsored videos so we try and keep everything as fair as possible. As these tools continue to leap frog each other, or long time subscribers know that we've had criticisms for each product. Your notion about the bias and something skewing our reporting is incorrect. Either way, we appreciate your engagement and thank you for watching!
I was able to drag and drop a jpeg into the Gen3 prompt area (Mac OS/Chrome). Nothing happened but then I wrote the prompt to animate the image and it worked.
It was the Gen 3 interface. Not Gen 2 for sure. It was the text to video prompt window. When the image was dropped an [Option] box appeared before the prompt. I did a few images to video but this seems to have been disabled now.
Runway 3 is SHIT value!! 100 credits for a video? Ridiculous. An absolute scam update, i burned through my entire month subscription with 6 generations before I realised the exorbitant price.
are you dumb? you should read the plans before buying. buying the cheap plans are a total waste of money. you should only go with the 99$ unlimited plan.
Videos are too expensive in gen 3 wow 100 credits. Its impressive though I tried doing an avalanche coming from a mountain. It looked photorealistic but the physics and everything else was just not physically possible if you know what I mean.
Hey Fella, I love your work, and what you bring to the comm. Thank you for making more compositions that are positive? Thank you for making more stuff that's spiritually progressive? It's so cool how we can make stuff that leaves the viewer feeling happy/. Like...... "Happy"....... Always digging deeper with our magnificent tools. That being said, what are the sys requirements to have a comm-independent type model?
thanks so much for the kind words! Definitely check out comfyUI if you want to learn more about having it on your own computer. To be honest, to make it an effiicient workflow you do need a good computer.
@@curiousrefuge As long as they have a PAID subscription, not those who have subscribed with a free plan. Gen-3 Alpha is available to those who have a paid subscription, Standard Plan + .
As a long time Runway subscriber, I have been waiting for improvements given all the new ai video alternatives. Unfortunately, while this is an improvement over gen2, it still has noticeable artifacts and disfigurement. Worst of all they increased the cost of using this "upgrade". Likely be dropping my subscription for free and less expensive alternatives
I make Film/Tv/Ads for a living - full time - it's been a very good career - a few name biggies etc.. own a great house/cars - survived the strike (just barely) - My company has great equipment, blah blah.. Now this AI video is interesting - but not really more than a novelty.. Good for very brief, limited transitions, establishing shots - maybe a flash of a close up for a non-recurring character/prop.. But for a number of reasons, AI Video doesn't deliver anything I can use for paying clients or industry standard deliverables.. I wish it did.. I really do.. but a number a upgrades are needed to make this a real tool we can use.. Maybe soon.. Thank you for your reports/vids - very good stuff..!!! I'll keep tuning in..
That should be obvious, my friend. Not to be rude, but no one has claimed these AI tools are actually good enough to replace industry standards. However, a contradiction to your statement, there are use cases of the tools being marketable. AI Generated or Assisted Trailers and Concept Films - This is the most notable use case for these tools. Proofs of Concepts, pitches, you name it. The Ai Film tools we are able to use are very able to serve needs in this aspect. People are also going into these projects knowing and understanding that the proofs are made with AI and respect that. A Trailer played very recently and showcased in Korea that was made in 9 hours has gotten the attention of potential interested parties in green-lighting an official production project. There may not be a lot of successes to this extent circulating at the moment, but they are there and they are growing.
I really think that this concept will not work. I am a very optimistic person and have a lot of video-editing experience. What need to be done is having a digital scene director for every scene (take). You should be able to use simple vector models to work fast. AI itself is great but can do too much. Too much means chaos and that' s right now happening. Prompts should be illustrators and not directors.
We understand the perspective, but the person *behind* the prompt should be considered the director :). Just as a director on a film set gives directions to the team, as we can give directions (prompt) to a tool.
actually i found you really DONT need to always use a specific format. its helpful but sometimes you can get results with simple prompts too like "a cat": it just varies.
You guys just amaze me - Runway Gen 3 blows it on every trial and yet you won't admit it. It totally ignores your prompts then if you happen to get a random great shot, you praise it to the skies. What gives? Why do you guys want to put your trusting subscribers through this incredibly disappointing experience?
There were reports that when this first got released, the servers were pretty cramped and so lots of people got unusual renders. This should be better now. What's your experience? We did not experience what you experience in our early access.
All I noticed after the big crush was that, rather than fixing the problem, Runway put out a "prompt guide" as if to say, "It's all your fault!" Also why isn't anyone talking about their absurd pricing structure? Do the math. Just figure out how far your credits will go and weep.
for what they charging i dont think the 15 dollar plan is worth it, i can see people looking messed up even in this video, with the guy with the fireworks. they have way to much morphing going on, runway is fun but for being usable , no, if i make a video i want it to be close to realistic, this is meh, still not there with the tech. u could get some usable stuff, but for all the prompts u will need to get something usable u will need to get the 95 dollar a month unlimited luma plan
Tried Gen2, didn’t like it at all. Trying Gen3, following the prompt structure and almost through 1000 credits and not one useful video.. the hallucinations are rough and even the interpretations of the prompts are not even close so far
You guessed it correctly! Nice! :) Email us with your name and youtube account here and let us know that you are the winner of the Game of the Week to claim your prize :) hello@curiousrefuge.com
"It's not perfect" => It's not good! Unlike AI generated images, which you can touch up in photoshop, there's no fixing the outputs here as far as I can tell. Of those 4th of July videos, none held up like the demos showcased at the start.
True - Text2Video outputs are very hard to clean up by itself. However, not impossible and typically requires rotoscoping/maksing (which you can use in Runway)
@@businessmanager7670I would but we can't since it's vaporware, lol. OpenAI also said it takes hours to render a few seconds. No way they're ever going to open that up to the general public. The queues would be months long.
I just finished a k-pop music video using both Gen-3 and Dream Machine. They're both worth it. Gen-3 was best at realism and resolution, Dream Machine, for dynamic motion. They are, hands down the best AI video platforms available IMO. There is still a lot of hallucination but for music videos all you're looking for are brief moments. So I consider this video the best thing I've done so far.
I canceled my membership months ago when I realized how little you get from the program compared to the price. I understand AI video has come a long way and from the looks of it, nothing will compare to Sora. However, I can’t justify paying $95/month just to have unlimited generations that take forever and end up having terrible outputs. Little to no movement, tons of blurry generations with deformities and mutations, and text input doesn’t really add much value. The many parameters don’t add much value either. It’s really providing a clearer (once you’ve done enough generations) slightly moving picture compared to other platforms. Not worth the hype at all.
Totally understand that there is a lot of cost to this and we're sorry to hear about your experience. The tools are advancing rapidly so perhaps it will make sense to check back again when it passes your threshold.
Can I use my own music and create my own prompts. f.e. "a poor 9 year old girl walking in the sunny park in a van Gogh style" on the beats of my own music?
@@curiousrefuge Sorry, but you did not read my question. I asked: "Can I use my own music ....". But Suno and Udio are music generators. I just do not want to use AI generated music. Cause that's how the music industry will die .
I am cancelling my runway account because they are insane if they think they should get money for Gen2 or Gen3. Luma, Haiper, and Pixverse are all better and I haven't paid a dime for them... yet. Even Pika is better.
They won't let Gen 2 be an option anymore, I'm ending my PAID subscription. I can't keep paying for videos before I see a picture of what I'm paying for, & no way to use Gen-2 again which had a preview option. You went too far Runway
No it's not. I've tested extensively for a real working production (animatic) and has TERRIBLE issues at understanding very simple tasks. Won't recommend investing in it unless for personal experiments.
Runway Gen-3 have priced themselves out of the Generative AI Video market. You can easily burn through a month's of credits in 20 minutes time....with pretty bad incoherent wonky results.
the mistake you make is thinking that we must create what has been done before . and failing to see people who are using AI in creative ways. I wrote a book on Claude hack.
In the trailer competition, based on who can something that looks holly hood. Forgetting that should really be the most creative use of AI in the genre of storytelling. Because I submitted a 10 minute story embedded within my trailer. I created consistency in each frame, something that AI is not good at. So perhaps the time has come to show the submissions to your public, I am sure even if you ask the wolve man(pulp fiction) he would say that is a good idea. So have a segment on your channel, that respects the artist who contributed to the competition, by watching and commenting on the videos a kinda, reaction to trailers in competition, I think that will make for interesting entertainment. Give them use of your platform for their creativity...
Ai hype! Is it too late to cease Ai? Remember Hansel and Gretal and that scenario? Luring us in with sweets? Just to be… laid off by Ai, then… suffer human extinction? Or worse, imprisoned by an… ai new world order? Enforced by swell robotics popping up everywhere… replacing you with Ai jobloss?
Out of 8 videos none of mine useable, all had odd morphing. The demos they are most main video creators showcase are taking their handful of best out of 100s maybe 1000s of rolls
Probably hundreds. But if anyone is a seasoned vet of using AI, they should know these are A - Cherry Picked and B - they should expect to run a dozen or more generations to hit a desired result.
@@kaelkent2496 No, it is not unusable. You get a refresh every week. The only limit you actually have to deal with is how frequently you generate. Space it out over the week and use your generations efficiently.
It is not hard...
and it is not unusable...
Maybe just be less picky and put the work in like the rest of us.
@@mythicproductions9154 seems like u focused only on the word unusable which was one part of my comment, i said "UNUSABLE ON THE 15 DOLLAR PLAN", and i stand by that , you need the unlimited plan for it to be usable. before calling me picky how about u read the whole comment. Gee wiz the 15 dollar plan you can easily run through 8 generations because of the small credits u get , so next time before you call someone picky how about you read the whole comment. Anyone considering doing this professionally by the way don't have time to wait a week for regeneration, no thank you, i would just get the 95 dollar plan if i want to WORK as you say. i work in video editing bud, how about you get off your high horse because you sound very condescending. The 15 dollar plan is a guinea pig plan for this company, and a waste of money, u need the 95 dollar plan for this to be worth it period. I've been doing ai several years now, i have made professional projects with it at work and compared to other ai i have used thats more consistent, yea the 15 dollar plan is not worth it.
@@mythicproductions9154 don't know why my comment disappeared but, no its not completely unusable, i said the 15 dollar plan is unusable, from the perspective of someone who works in editing, only the 95 dollar plan would be worth it, no one who works professionally would want to wait a week for regeneration of credits on deadlines, and to say im picky how about you get over your assumptions and condescending attitude when u didn't take the time to read my comment and focused on only the word unusable, i didnt say its completely unusable, i said the 15 dollar plan is unusable also people like u who don't advocate for better than subpar and are happy wasting 15 dollars for a subpar product will allow companies to keep taking advantage of you. People kept braggin on adobe and then now all of a sudden that they see the price model and the terms of service that they can train on your work in adobe now want to fight for rights when it should have been done long ago. I will advocate for improvement that this is not acceptable on a 15 dollar plan.
i agree with you bro, its not worth it on the 15 plan its unusable. the 95 dollar plan is the only plan that's usable for something professional with deadlines.
IN REGARDS TO THE user mythicproductions9154
The $15 plan is what I found unusable. As someone who works in editing, only the $95 plan is worth it. Professionals can't wait a week for credit regeneration on tight deadlines. Your assumption about me being picky is misplaced. Instead of focusing on just the word 'unusable,' consider reading my comment thoroughly. Accepting subpar products and wasting $15 allows companies to take advantage of you. Advocacy for better services is crucial. I stand by my point that the $15 plan needs improvement.
Thank you for your honest feedback on Runway Gen 3. I also really like how you often have a segment that compares and contrasts different AI software
Our pleasure!
Gen-3 is pretty amazing. Content moderation is a bit hyper sensitive. It blocked this prompt a couple of times. I think because I references "exposed" wiring. Here is the prompt. It was part of a sci-fi horror short version 0.1 I posted using gen-3 where I am going to show my progress and improvements on the short. Prompt: " Wide-angle tracking shot: The camera sweeps through a damaged hallway in a futuristic space station. Exposed wiring sparks, with fire and smoke erupting from wall breaches. Automatic doors glitch, unable to fully open or close. The scene is illuminated by erratic emergency lighting." They will get that part right I hope.
I think the AI just rejected your script on the grounds of taste?
I was able to visualize it really easy while reading it. well written for humans. hope gen3 gets it. - if "exposed" is causing you trouble maybe try other wording for just that part
@@animalaliens I shared this and some others that were blocked. They agreed they should not have been and are fine tuning the moderation. Safety first and I agree but it is still in Alpha state and they will get it right at some point.
True! You have to be careful with the words - runway was already pretty strict on prompts prior to Gen3
I made a music video with Gen-3 and it cost me about $30 in credits. As a music producer and artist using it in the real world, I would say it's worth it.
"There's a sucker born every minute" - P. T. Barnum
Having used a similar amount of credits, i have to say I question your judgement. It is a novelty atm. If people watch your music video 10 years from now, what do you think they will think of it? How well do you think they will think it has aged?
@@bjorn0helander u can watch the video and make your own judgement, I’m just tryna be creative with the tools at the moment
@@Kaotonix Watched it (assuming it is the one with the waves and the fish). I still don't think it's a worthwhile tool at the moment. Too bad prompt comprehension, too chaotic (not your video, the tool - I can see you did enough editing to temper that). For now, I would rather use something like Luma, or just run stable diffussion locally. I couldn't really recommend runway v3 as it is. Image quality is great though. If they allow img to vid, may be different. Anyway, that video was better than I expected, and music not AI generated (thought it would be someone who "made" a track in suno), so props for that.
@@bjorn0helander thanks for taking the time, yes I’ve been making music for a decade and never really had a great tool for visualizing the sonic soundscapes I spent so much time crafting. I def spent a lot more money on previous projects so I knew what I wanted with this one and had experience prompting, which is how it cost me $30. Honestly I tried luma and others, including gen 2. Def did not yield anything worthwhile. For me, gen 3 was the first time it felt useful. Anyway, cheers 🥂 we have an exciting future for video generation. I tried stable diffusion but my graphics card is shit so didn’t work out for me
The *keyframe* tool in *Luma* would be much more useful if they add these *2 features:*
*1.* Ability to add multiple keyframes. I want to add more than 2 images, because an action may have multiple steps to it.
*2.* Transition between frames using camera effects. I want to be able to choose a camera effect to apply when transitioning from one frame to another.
I can give you a tip for that. First you use your 1st 2 images. when you want to download your animation, you will see that you can also download just 1 image if you want. So you could download just the last image of your animation to calculate a new animation with that image as Input and you can add an other image as output image. It works very well
you dont need more keyframes in a 5 seconds video
Multiple frame would be amazing. Imagine all of the interesting shots we could put together!
Layers and exr files = essential
Control = Essential
The input should be a 3D app and the final out output should be a composite.
The goal should be to create render layers for a final result
We're excited for there to be far more control in future updates!
I made a music video with Gen 3. While using it this can be very frustrating experience but after some time you learn the limitations of Gen 3 and get some nice results which are better than Luma or Haiper which I used before. You also need the unlimited 95$ plan.
Fair tip! It requires lots of rolls and rerolls!
Great work, Curious Refuge. Keep it coming please.
Thanks, will do!
I just made my first AI film using Runway Gen-3 after trying Luma. Wasn't sure I could handle being limited to text prompts after seeing how amazing Luma is at turning images to video. But once I got used to Gen-3 prompting, I actually liked it better. You get to be a movie director rather than just trying to get images to do something interesting (and mostly failing at it).
Glad Gen3 worked out for you!
@curiousrefuge, which AI screenwriting tool are you referring to at the 00:09 mark?
Claude :)
I have some establishing shots that I used in Gen-3 and was impressed how downtown Tokyo looks at midnight. I wish I could upload pics tho and I would like to pick my aspect ratio.
True! We can't wait for I2V to be available!
A lot of this looks to be rendered in slow motion. Does anyone know the default FPS (frames per second) and resolution? I realize that version Alpha has limited control and functionality, but much of the action in the background is really unusable unless the application is comedy, at least in the examples you've shown here.
i believe it is intentional . there should be a problem with integrating physics and and angles in real time . they hide this glitch by going with slow motion.
For now most of AI tools render in slow motion. However, these tools are getting better each month and Luma already has some quite fast renders.
I'm extremely disappointed with the pricing and quality of Runway Gen-3. The cost is high, especially considering the wasted generations and poor physics, which make it almost worthless for the average consumer. This product shouldn't cost more than Midjourney on a monthly plan given its current issues. The price doesn't reflect the quality at all. Unless you're a business, it's hard to justify the expense. The pricing structure is a major letdown.
Are you forgetting that price is mostly decided by their cost to run the service at this point? You're demanding the impossible.. Reminds me of the "angry man shouting at the sky" meme
@@pictzone More like the old dehydrated dying man in the hot desert sands and the bottle of water is just out of reach. If these tools cost just as much as traditional filmmaking, why would anyone use these inferior tools. AI was supposed to empower the everyman, not just the rich production houses flush with cash. At least Luma gives us 30 free generations per month. Runway needs to throw people a bone.
@@pictzone Dumb.
They're complaining about value (which is valid). Value has nothing to do with price.
Why pay 10 times more for something that is twice as good.
@@VJ4rawr2 So they should finance small creators out of their pocket right? Have some patience, this tech will be affordable in less than 6 months. Be thankful it's "relatively" afforadble right now, although with limited control on the creative output. The 100$ package is very generous for what it provides in my opinion (check it out if you don't know what I'm reffering to)
It's true - these tools (especially when they come out) have quite a pricey tag. However, with more competition these prices will change (or else a competitior will take the market). These tools are still very new so let's just keep an eye out. We certainly understand the prices hurting the average users at this point!
I must admit that I prefere Luma than Ruway for the moment and it is cheaper, and it gives very good results with Input and Output pictures very quickly.
Fair point! Image2Video is crucial for AI filmmaking!
Does luma allow you to stop the camera from moving/panning on it's videos?
@@starterplanet its hit or miss. you put it in the prompt. static shot helps.
No preview button on Gen-3 before the high cost of vid generation. I was a Runway fan, will be switching to something better now.
A preview button would be amazing. I'm sure they'll have it soon!
@@curiousrefuge They had a pic preview on Gen-2
its TOTALLY WORTH IT on the unlimited plan, on credits alone no it may not be but i would HIGHLY encourage anyone interested to get unlimited for the month the renders are a tad slower but you can play around as much as you want and to me the quality beats anything else right now you can get the prompts down pat but it may take more then one generation to get the output you want so thats why again i say unlimited is CRITICAL for anyone wanting to use this unless you just have a ton of money to throw at it, otherwise you WILL burn through credits FAST for sure!
That's exactly what I'm thinking about it. 👍
Oh baby is it ever!
Can I borrow $100?
I have one question? Can it keep faces consistent in different prompts so it can be used to create short films and all
We agree however that initial cost is quite expensive!
Did Runway AI's gen 3 updated again? on their pricing modle the lowest paid plan says "Gen-3 Alpha Turbo (Image to Video) up to 10 sec"??? Or i am guessing it doesn't have image to video for you because maybe it was the free version you were using?
We use the unlimited version!
I really like your videos. But I am quite lost in it. Pls, what AI platform could work with a real video then parexample to keep a real persons of the video then recreate a background or only create an artistic things abound those real persons? I mean in a very good quality. Thank you very much.
I'd check out Runway if I were you!
@@curiousrefuge I did. It is a sh.t, a fraud. One generates a video with no sure results but payed ofcourse, it doesn 't follow the prompts at all, even the easy ones. It is a joke.
Super excited about Gen-3, really hope it catches up to Luma soon.
We love the competition!
I tested the LTX studio Beta and their character consistency is a real game changer
I signed up for the waitlist but am still waiting. Did it take you long to be able to use it after signing up?
We believe that is definitely a powerful tool!
@@huzailijamal117 dunno. Got a content creator website. Maybe it helps
please review LTX studio. shot 1: Luma
shot 2: Runway gen-3
In a previous video we gave it a quick shootout but we'll consider a longer version :)
Runway gen 3 is worth much more than the old versions
fair point!
What screen writing software are you talking about at 11 seconds in this video?
Claude 3.5
What was the screenwriting software you mentioned at the beginning?
Fantastic summary - great collection of the state of the art.
Thanks for watching!
Thank you for your honesty. I can sleep calm. My job is safe for many more months by the looks of it..
You are so welcome XD. Glad you enjoyed this video :)
Fantastic update and review geared to filmmakers
So glad you liked it!
Yes ... on fire here in Texas Summer - and we say " Nuculear" like it sounded 7:17
Most of our team is from Texas :)
shot 1: Luma
shot 2: Runway gen-3
Given the results achieved, this is my idea, they shouldn't ask for money. When they have reached a decent quality, then it is right that they ask you to pay.
So... who is paying for the servers to run and generate everyone's content? Money does not grow on trees and Runway is not exactly as rich as Elon Musk.
Second, if you are familiar with using Ai at all, it goes without saying that you HAVE to cherry pick for the best results. It is a part of the process.
@@mythicproductions9154 I know well that the company that manages the service has to pay for the use of the servers, the programmers...etc. But there is also something called: business risk. An alternative could be crowdfunding, which seems more coherent to me for this type of initiative.
Given the needs to running a business this probably wouldn't happen.
Hi CURIOUS REFUGE; Nice vidéo, do you think than on RUNWAY will can put IMAGE TO VIDEO like Luma or other ( haiper,...)
Yes we believe it will soon!
Is there any indication that if you get a result with bizarre artifacts or physical impossibilities, you can re-roll more times, , adjust the prompt, and finally get it to look completely realistic?
Yea... forgive me but that is how prompting works. From the first generation and onward you are either tweaking the prompt or re-rolling results. How much you need to do either varies. I have personally taken on a workflow where I prioritize based on the following: Art Style -> Visual Tone -> Setting Description -> Included elements (People, items, etc).
Video it can be a bit harder but I run with the same practice in mind. Another habit I started is before I prompt for a video (if I can't or don't have ref images), I will use the prompt in an image generator to get an idea of the results I could expect, dome some tweaking, then take it to Runway. If I can get what I want well enough in images, I will use a keyframe tool like KREA to generate the video instead.
It's all about rolling and re-rolling :)
So are we talking about luma and Claude to compare what Runway can't do or can it do the same. I am confused.
Yes we compare and give our analysis. Would you mind elaborating a bit more on your question?
would be nice to read the exact prompts on the generated videos, thank you!
Noted! We'll try and add that info for next time :)
Are you in front of a green screen or a projector?
A projector in the room! We are testing it out. What do you think?
@@curiousrefuge It looks awesome! :)
Great show with super info.
Glad you enjoyed it!
People who write about making videos on Runway, can you put links to them in the comments?
Sure!
I tried the tool straight away because the examples look fantastic.
The results were more than disappointing. At first I thought it was the prompts.
Then I tried a few original prompts from the website. Unfortunately, they rarely produced the same result.
They really need to improve that.
We've heard similar stories from other users. It seems it could be do to the extreme use by lots of people at the same time (stress on servers?). Not sure if this is true, but it seems like people are having much better luck now!
@@curiousrefuge Thank you for the information. I will try it
First video clip is created by Runway and the second Luma 😊
Paid subs on Luma take 2-4 minutes and since there is no free version in Runway (yes, the free version of Luma can be quite slow), how about an apples to apples comparison and ditch the bias? That's a pretty big detail to miss so what else is being skewed in your reporting?
Runway does have a non-paid tier. About 130 seconds to be specific. It is not amazing but it is not bad. And it is common practice with most AI tools to reserve their latest releases to their paid clients first before going full problem. Open AI operated like this. Runway Operates like this. Most do. These are well known aspects of the AI space sense last year if not earlier. He did an apples to apples comparison to the capabilities of both products. If you don't like that he didn't review business model details, grow up, stop being lazy, and go check out the platforms yourself and read. This subject was touched on when Runway first released a long time ago and their business model has been the same sense. There is no reason to review it again without a worth while notable update.
@@mythicproductions9154 Not for Gen 3 as far as I can see. (only for gen 2). The issue is: Caleb is implying that while Runway's paid tier is fast, Luma's is slow and that couldn't be further from the truth. Luma's free tier is slow, but Runway Gen3 does not have a free tier to compare that too (I wonder why that is? Would it too be slow?)
@@markrichards5630 Well, yes and as I pointed out it is standard practice for a business like Runway to roll out these updates to paid users before they are open to general free use.
Idk how long that will be but that has been the practice in the past and that is what I expect going forward.
I can not comment to the gen speed of Luma, but other factors could play into this. High user volume will slow a server down. If he is hitting long delays in generations, it could be because of user volume, internet connection, interfere, and other issues I am not sure I am thinking of.
My only problem with the OG comment is the jab and assumption that what was covered in the video is with bad intent or at least that is how it sounds. Also, might I add, it is borderline impossible to expect someone like him in his place to not be biased on a platform.
Regardless of what a content creator may say, it should go without saying everyone should do their own research and make the decision for themselves which platform they prefer.
We may be the only channel (or relative to our size) that doesn't take sponsored videos so we try and keep everything as fair as possible. As these tools continue to leap frog each other, or long time subscribers know that we've had criticisms for each product. Your notion about the bias and something skewing our reporting is incorrect. Either way, we appreciate your engagement and thank you for watching!
I was able to drag and drop a jpeg into the Gen3 prompt area (Mac OS/Chrome). Nothing happened but then I wrote the prompt to animate the image and it worked.
You were using gen 2 not 3
This sounds like Gen2 :)
It was the Gen 3 interface. Not Gen 2 for sure. It was the text to video prompt window. When the image was dropped an [Option] box appeared before the prompt. I did a few images to video but this seems to have been disabled now.
Runway 3 is SHIT value!!
100 credits for a video? Ridiculous.
An absolute scam update, i burned through my entire month subscription with 6 generations before I realised the exorbitant price.
are you dumb? you should read the plans before buying. buying the cheap plans are a total waste of money. you should only go with the 99$ unlimited plan.
True, quite expensive!
Videos are too expensive in gen 3 wow 100 credits. Its impressive though I tried doing an avalanche coming from a mountain. It looked photorealistic but the physics and everything else was just not physically possible if you know what I mean.
True! We hope the costs will come down soon!
luma is way cheaper. like 10x the number of vids for the same price.
Shot 2 is Luma
Shot 1 is Runway.
It's pretty clear that Runway's videos have a more drawn/cartoonish look while Luma leans towards more photorealism
Hey Fella, I love your work, and what you bring to the comm.
Thank you for making more compositions that are positive?
Thank you for making more stuff that's spiritually progressive?
It's so cool how we can make stuff that leaves the viewer feeling happy/. Like...... "Happy".......
Always digging deeper with our magnificent tools.
That being said, what are the sys requirements to have a comm-independent type model?
thanks so much for the kind words! Definitely check out comfyUI if you want to learn more about having it on your own computer. To be honest, to make it an effiicient workflow you do need a good computer.
Gen 3 is currently an "Upgrade to try" situation, right?
Available to anyone so long as they have subscription.
@@curiousrefuge As long as they have a PAID subscription, not those who have subscribed with a free plan. Gen-3 Alpha is available to those who have a paid subscription, Standard Plan + .
Keep crushing guys! Still the best source for info and inspiration on UA-cam!
Appreciate it!
As a long time Runway subscriber, I have been waiting for improvements given all the new ai video alternatives. Unfortunately, while this is an improvement over gen2, it still has noticeable artifacts and disfigurement. Worst of all they increased the cost of using this "upgrade". Likely be dropping my subscription for free and less expensive alternatives
Totally fair points! I'm sure they will make some upgrades soon!
O this is what I’m looking for time to make weather movies
Glad you found this helpful!
I make Film/Tv/Ads for a living - full time - it's been a very good career - a few name biggies etc.. own a great house/cars - survived the strike (just barely) - My company has great equipment, blah blah.. Now this AI video is interesting - but not really more than a novelty.. Good for very brief, limited transitions, establishing shots - maybe a flash of a close up for a non-recurring character/prop.. But for a number of reasons, AI Video doesn't deliver anything I can use for paying clients or industry standard deliverables.. I wish it did.. I really do.. but a number a upgrades are needed to make this a real tool we can use.. Maybe soon.. Thank you for your reports/vids - very good stuff..!!! I'll keep tuning in..
That should be obvious, my friend. Not to be rude, but no one has claimed these AI tools are actually good enough to replace industry standards.
However, a contradiction to your statement, there are use cases of the tools being marketable.
AI Generated or Assisted Trailers and Concept Films - This is the most notable use case for these tools. Proofs of Concepts, pitches, you name it. The Ai Film tools we are able to use are very able to serve needs in this aspect. People are also going into these projects knowing and understanding that the proofs are made with AI and respect that. A Trailer played very recently and showcased in Korea that was made in 9 hours has gotten the attention of potential interested parties in green-lighting an official production project. There may not be a lot of successes to this extent circulating at the moment, but they are there and they are growing.
We agree! Caleb often says these tools are not quite ready yet for the silver screen. However, in just 1 year we've come a long way!
I really think that this concept will not work. I am a very optimistic person and have a lot of video-editing experience. What need to be done is having a digital scene director for every scene (take). You should be able to use simple vector models to work fast. AI itself is great but can do too much. Too much means chaos and that' s right now happening. Prompts should be illustrators and not directors.
We understand the perspective, but the person *behind* the prompt should be considered the director :). Just as a director on a film set gives directions to the team, as we can give directions (prompt) to a tool.
1. Runway 2. Luma
(based on image detail and coherence)
Sparkler is two syllables. Not sparkle-er. :) Otherwise, good video.
Hehe, we appreciate the lesson ;)
actually i found you really DONT need to always use a specific format. its helpful but sometimes you can get results with simple prompts too like "a cat": it just varies.
Fair point!
So it can only do hyperrealism? Can’t it do semi-realism and anime-style cartoons? When is ‘Image to Video’ coming to Runway Gen-3? Did they say?
You can try any of them! We expect I2V will be soon but we don't have a date for it.
I expect Gen3 videos will flood the internet. Its good enough
It's certainly already flooding the social feeds!
I'm laughing my a** of with all the chaos. When Ai looks at us, this is what it sees?? lol
It's certainly a beautiful mess :)
5:53 is all anyone needed to know
Thanks for tuning in!
The way it follows prompts, I call it Runaway Gen 3.
I see what you did there :)
Every result has pinch of "Horror" touch to it 🤣🤣
Very spooky :)
You guys just amaze me - Runway Gen 3 blows it on every trial and yet you won't admit it. It totally ignores your prompts then if you happen to get a random great shot, you praise it to the skies. What gives? Why do you guys want to put your trusting subscribers through this incredibly disappointing experience?
There were reports that when this first got released, the servers were pretty cramped and so lots of people got unusual renders. This should be better now. What's your experience? We did not experience what you experience in our early access.
All I noticed after the big crush was that, rather than fixing the problem, Runway put out a "prompt guide" as if to say, "It's all your fault!" Also why isn't anyone talking about their absurd pricing structure? Do the math. Just figure out how far your credits will go and weep.
for what they charging i dont think the 15 dollar plan is worth it, i can see people looking messed up even in this video, with the guy with the fireworks. they have way to much morphing going on, runway is fun but for being usable , no, if i make a video i want it to be close to realistic, this is meh, still not there with the tech. u could get some usable stuff, but for all the prompts u will need to get something usable u will need to get the 95 dollar a month unlimited luma plan
Many others are voicing their tip that the $95 plan is the best to use. Fair point!
Tried Gen2, didn’t like it at all. Trying Gen3, following the prompt structure and almost through 1000 credits and not one useful video.. the hallucinations are rough and even the interpretations of the prompts are not even close so far
Much agreed. The morphing characters are unhinged
Fair! Many people have said the cost structure just isn't quite there yet with the number of re rolls.
Shot 2 - Luma | Shot 1 - Runway Gen 3
9:40 1: LUMA 2: RUNWAY I guess
You guessed it correctly! Nice! :) Email us with your name and youtube account here and let us know that you are the winner of the Game of the Week to claim your prize :) hello@curiousrefuge.com
@@curiousrefuge wow, ty!
Very interesting
Glad you think so!
What could Runaway Gen-4 be like?
Will be craaaaazy!
"It's not perfect" => It's not good! Unlike AI generated images, which you can touch up in photoshop, there's no fixing the outputs here as far as I can tell. Of those 4th of July videos, none held up like the demos showcased at the start.
True - Text2Video outputs are very hard to clean up by itself. However, not impossible and typically requires rotoscoping/maksing (which you can use in Runway)
cool vid
Glad you enjoyed it!
I hope that SORA AI is released sooner rather than later...because Runway Gen 3 and Luma both really blow and look amateurish at best!
SORA doesn't look any better. Worse, actually.
@@High-Tech-Geek false sora looks way better (like camera footage) and has more consistency. stop spreading bs or provide evidence.
@@businessmanager7670I would but we can't since it's vaporware, lol.
OpenAI also said it takes hours to render a few seconds. No way they're ever going to open that up to the general public. The queues would be months long.
We don't think it will come out anytime soon! At least a few more months...
1.Runaway 2. Luma
Game :
Dog clip 1 Luma, dog clip 2 Runway Gen 3
I just finished a k-pop music video using both Gen-3 and Dream Machine. They're both worth it. Gen-3 was best at realism and resolution, Dream Machine, for dynamic motion. They are, hands down the best AI video platforms available IMO. There is still a lot of hallucination but for music videos all you're looking for are brief moments. So I consider this video the best thing I've done so far.
Is 'hallucination' actually a word lots of people are using to describe f**ked up bit of AI renders?
@@Hammy5641 Yeah it used for both generative and language models when they start making up things that don’t exist in reality
We appreciate you using a balanced workflow! Getting the best out of both worlds :)
I canceled my membership months ago when I realized how little you get from the program compared to the price. I understand AI video has come a long way and from the looks of it, nothing will compare to Sora. However, I can’t justify paying $95/month just to have unlimited generations that take forever and end up having terrible outputs.
Little to no movement, tons of blurry generations with deformities and mutations, and text input doesn’t really add much value. The many parameters don’t add much value either. It’s really providing a clearer (once you’ve done enough generations) slightly moving picture compared to other platforms.
Not worth the hype at all.
Totally understand that there is a lot of cost to this and we're sorry to hear about your experience. The tools are advancing rapidly so perhaps it will make sense to check back again when it passes your threshold.
For me wasn’t good enough. Still neeed more credits each charge to make things okay.
Fair point! Still takes lots of rolling and rerolling !
1.) Luma
2.) Runway Gen 3
1. Runway Gen 3
2. Luma
Shot 1 is Luma
Shot 2 is Runway.
1. runway 2. luma
Can I use my own music and create my own prompts. f.e. "a poor 9 year old girl walking in the sunny park in a van Gogh style" on the beats of my own music?
You can test it out in Suno or Udio!
@@curiousrefuge Thanks from Lemmer (The Netherlands) I'll try🙋♂
@@curiousrefuge Sorry, but you did not read my question. I asked: "Can I use my own music ....". But Suno and Udio are music generators. I just do not want to use AI generated music. Cause that's how the music industry will die .
not so useful but works for stablishing shots, drone stuff. still it generates a lot of useless stuff.
True! Great for establishing shots!
1- Luma 2. Runway
1. Luma 2. Runway
Shot 1 - Luma
Shot 2 - Runway
Interesting observation: the burgers must be imaginary. They can't eat them and they phased through them xD
1. Luma
2. Runway Gen-3
I am cancelling my runway account because they are insane if they think they should get money for Gen2 or Gen3. Luma, Haiper, and Pixverse are all better and I haven't paid a dime for them... yet. Even Pika is better.
those are not better lol. those are trash, not sure what you are smoking.
Luma also costs 400$ per month and is not unlimited
We agree that the prices are quite expensive. We hope these will come down very soon!
They won't let Gen 2 be an option anymore, I'm ending my PAID subscription. I can't keep paying for videos before I see a picture of what I'm paying for, & no way to use Gen-2 again which had a preview option. You went too far Runway
No it's not. I've tested extensively for a real working production (animatic) and has TERRIBLE issues at understanding very simple tasks. Won't recommend investing in it unless for personal experiments.
We appreciate you sharing your experience!
Runway Gen-3 have priced themselves out of the Generative AI Video market. You can easily burn through a month's of credits in 20 minutes time....with pretty bad incoherent wonky results.
True! We hope the prices change soon!
Shot 1 Luma
Shot 2 gen3
Terrifying uncanny valley
We're definitely in that realm! Once you get use to the tools it doesn't feel so terrifying though :)
the mistake you make is thinking that we must create what has been done before . and failing to see people who are using AI in creative ways. I wrote a book on Claude hack.
We disagree. We believe AI is a great tool to remake, remix, and/or create entirely new ideas.
@@curiousrefuge That is my point, look for creative ways people are using AI, instead of focusing on trying to do what was done before.
In the trailer competition, based on who can something that looks holly hood. Forgetting that should really be the most creative use of AI in the genre of storytelling. Because I submitted a 10 minute story embedded within my trailer. I created consistency in each frame, something that AI is not good at. So perhaps the time has come to show the submissions to your public, I am sure even if you ask the wolve man(pulp fiction) he would say that is a good idea. So have a segment on your channel, that respects the artist who contributed to the competition, by watching and commenting on the videos a kinda, reaction to trailers in competition, I think that will make for interesting entertainment. Give them use of your platform for their creativity...
Luma is still more impressive than Gen 3 for me
We hear ya! We love Luma too!
The dog 1 runway dog 2 luma
mr beast 5:36
Yezzir! We love Jimmy! :)
1: luma,
2: runway
Ai hype! Is it too late to cease Ai? Remember Hansel and Gretal and that scenario? Luring us in with sweets? Just to be… laid off by Ai, then… suffer human extinction? Or worse, imprisoned by an… ai new world order? Enforced by swell robotics popping up everywhere… replacing you with Ai jobloss?
We appreciate your engagement but we disagree with the premise :)
Ask it to make people dance, it's very durpy
Haha yes that would not work out well!
Just toys toys toys.... I'm not impressed yet.
Stay tuned ;)
Is it porn compatible? How their porn look like?
Not possible with these tools for safety reasons!
Good video, but OMFG please don't ever say "nukular" again. It's nuCLEAR (noo-klee-ur).
Haha that’s my Texas accent coming out.