Flora’s concept is certainly interesting! 🤔 Definitely a lot like a storyboarding/brainstorming program. I hope a local equivalent gets developed at some point!
@TheoreticallyMedia Tim you are more consistent and dependable than any professional athlete on the planet. One day during the future AI filmmaking Oscars, your jersey will go up in the rafters, statues will be built when we look back at your helpfulness and curiosity during this AI filmmaking renaissance. Been watching all year and wanted to stop and say THANK YOU! YOU DA MAN!
Every person ever talking about nodes : Oh it's not that complex... Then opens up a what looks like a map of the London underground with its 500 miles of rail network and 300 destinations 😭
Haha- I always call it spaghetti, but the london underground map is also pretty accurate! It looks complex on first glance, and then you think you've got it...only to realize: Wait, no-- it IS complex. haha. On the plus side, when you build out a node system, at least it's YOUR mess.
I'm not going to call out the YT'er I'm referring to, as I do like them, but they always say the same type of thing about Comfy - it's not hard once you know it and so on. Yet, whenever they go live with it they spend half the time sorting problems and having things not work. That says it all.
@Elwaves2925 it's right throughout ... I was learning unreal and joined online course theyd say the same about materials using nodes.... Always "its not that hard when you look into it" then spent 1 hour going over it, skipping through parts by saying "we'll cover that in a later date"...
Totally-- I think they're great options that serve different purposes. The fact that you can hone in quickly on what model is best for a style in Based is pretty great-- and Flora might take you into some interesting story concept directions.
In Flora, is that 15 minutes of just being on the site or 15 minutes of generation time or 15 minutes of video rendered out? I also just saw in their price plan it mentions "hours of realtime". What is that?
So, I don’t think Flora is good at that yet. Best bet there might be to generate/train a model in Krea, Based, or really anywhere that model training is popping up. And then import into Flora as an image source. I’ll reach out to them to suggest it!
Couple of things: 1. How are all these platforms offering other models all-in-one? I don't get it. Are they tamed-down versions of Flux, SD, Runway, Kling, etc.? Something doesn't feel right. 2. Would appreciate slightly more in-depth one with Flora. 👍 Thanks as always, Tim. 💯
Some of the commercial models are available to use via API, some of the opensource models could be deployed locally. With commercial models they usually just overcharge you (significantly) compared if you'd use the model on the original site - this is their profit. With open source models they charge for the used resources. Usually they not only overcharge you for resources but also use simplified versions of the models to generate faster and reduce resource usage. You'd never know what crippled version of Flux you use.
Base labs looks good but a lot of apps that I will never need I looking to start doing ai generated videos I like Kling flux runway and their new talking app but I also want model training face swapping upscale etc clothing swap would be a bonus. Going this route I would have sign up for 3-subscriptions would there be a way to bring it down to 2 subscriptions? My budget would be around 40 per month. thanks love the videos! Oh If this helps I use DaVinci resolve and premiere pro and I am pretty proficient at both.😂
I'm always thinking about switching to Davinci. I CAN use it-- I'm just really slow at it. Sadly, because of the nature of AI, and doing these videos, I never really have a chance to sit down and GET fast at Resolve. Final Cut ALMOST got me interested in this latest version-- but, yeah...still gotta pass on that one. (I'll say, the one killer feature for me in Premiere is still the Transcription Edit. For "talking head" videos like this-- I mean, it shaves HOURS out of my edit.)
@@TheoreticallyMedia lol but I still need to know which 2 sites would give me best bang for the buck to start ai video creation? In your experience what 2 sites would be best? Thanks Ps… I used premiere for for 15 years I now get it free from my sons edu subscription. But DaVinci is free or one time of 300. Forever! If you can force yourself to do 3 projects in a row you will be more comfortable with it. It also has transcription editing like premiere I use it a lot too. Bonus it has smooth cut so if you cut in the middle of talking head you will not see the cut. Premiere has something like it but it’s a bit wonky sometimes. Ai in it now has auto voice isolation and music mix editor so you can turn up or down you can isolate every instrument take voice out of a song, just leave the drums take out the rest of the band sure no problem.
NightCafe is similar to Based and has been around since the beginning. The problem I have with most platforms which offer a variety of models at your fingertips is the censorship. 🤬 At least with mystic ai access to Flux, there's minimal censorship 👍 👍 👍 Too bad mystic ai isn't a platform that offers the donut shop style selection of models that NightCafe offers. 😔
@@TheoreticallyMedia It's a donut shop where you can very conveniently select from a large variety of models. The NightCafe interface is excellent IMO
@TheoreticallyMedia NightCafe is still quite relevant in the AI space....... considering how old they are, this is quite a feat! They thrive in the community aspect which for those who need or love that, is great. I, as a secret misanthrope, am not enamored by that part of NightCafe 😉
@diane161061 And who doesn't like donut shops where you have a nice variety of different styles to easily choose from right at your fingertips! That's why I do like NightCafe....... it's like a donut shop.
These workflow GUI are kind of dumb in my opinion, They would easily be able to change this into a combo boxes with options or something similar for ease of end user and would appeal to a lot larger audience.
For me, the elephant in the room is lip sync. You either get good lip sync which lasts for longer than ten seconds with very little character motion, or good character motion with little to no lip syncing (other than maybe ten seconds or less). I'm waiting for that (Runways Act one is close).
Flora's capabilities look fun, but the murder-board does not 😅 I hope this stupid trend dies off, or, at least, they start providing more neurodivirgent-friendly organisational options (folders, trees, etc.) I already spend around $400/month on this hobby, so it is going to take more than another aggregator to get me to switch, or even try new stuff.
@TheoreticallyMedia Thanks, Tim! Means a lot, coming from the Dad of AI Dad Jokes 🤘❤️🙇♀️🇦🇺🐨🦘 (I am also kinds famous amongst my students for mum jokes - both of them).
Hey Weber here I’m the founder of FLORA - we’re working on more neurodivergent friendly organization options. We hope to add everything in one place in a new type of workflow-oriented UX so in addition to acccessing everything, you can stitch them together and automatically run a chain of FLUX > Image to Video nodes with different prompts at the same time, and share your projects with other people with real-time multiplayer. We’re just in Alpha so hopefully we justify over time our node-based approach :)
I find that none of these video models do black faces all that well. Well they do for some, just not mine. They always change my features... drastically
Nice video! But most of all-in-one platforms are so ridiculously overpriced compared to the pricings of the original tools they provide functionality of that anyone who managed to graduate the elementary school won't touch them with a 10-feet pole. And carry over credits made me chuckle. Something like "Here are 1000 credits for only 99.99, one generation on Flux cost 10 credits and one Kling generation cost 50 credits but you can carry over month". Yeah, good luck with that.
Eventually there will be a google one, that already has all my images sorted by faces, so it should be able to perfectly recreate me, friends and family and all the places I often photograph. Then I can say "ok google you know were we've been the last 10 days (city tour Paris) please make some photos and videos from me and my girlfriend that fit the places we've been to (time and weather correct) please."
Two tools I Should know about??? Really??? 14 free credits? $25 for 1,000 credits with a basic result. Seriously??? I hope they paid you a lot of very good money' Because you're probably familiar with the saying: Tell a lie once, and all your truths become questionable. I know this sounds a bit dramatic, but I feel betrayed. :(
Rumor is that Sora releases sometime this month. Literally just asked David from Midjourney if Video was shipping before the end of the year-- he said probably closer to Jan, only because the team doesn't want to work over the holidays (totally understand that).
Oh, these 2 AIs will help content creators have many ideas, saving a lot of time and effort. Thank you for sharing
Flora’s concept is certainly interesting! 🤔 Definitely a lot like a storyboarding/brainstorming program. I hope a local equivalent gets developed at some point!
Tim is our ai dad haha
Gonna need you to clean your room before I let you generate another image! haha
😂
More like our AI Fred Flimtstone. But with less body mass and better taste in clothing. The hair is about the same though. 😆
No cap
@@Apuat I think you meant "Fred Filmstone."
hahaha pop off in the clerb -- excellent video as always tim
haven't even heard of these so thanks for shouting out, will try them this weekend
This is awesome! Thanks for including us!
1000%! Looking forward to seeing how everything develops in the next few months!!
Flora looks like a lot of fun and BaseLabs a great all in one suite.
Always like before watching the full video :) !!! thanks for thsi video
appreciate that! as well as the comment!
@TheoreticallyMedia thanks you for this great work that you are doing.
Interesting worklow, I wish one could convey an idea in such a way, visualize it and perhaps develop it further in dialogue.
Amazing video as always, Tim. The commercial AI world is super exciting right now
Your subtle humor is excellent along with your "infotainment" style. You earned my sub :)
@@TheHorrySheetShow He would be a great person to have a drink with. Very funny.
@TheoreticallyMedia Tim you are more consistent and dependable than any professional athlete on the planet. One day during the future AI filmmaking Oscars, your jersey will go up in the rafters, statues will be built when we look back at your helpfulness and curiosity during this AI filmmaking renaissance. Been watching all year and wanted to stop and say THANK YOU! YOU DA MAN!
The most consistent character in AI Video well at least tied with Trisha Codes pink bob and glasses
A Return to Forever Easter egg! Nice. 👍🏾🎹
Any recommendations on the advanced settings for training?
get away from the angry dude with gun ! hahah... tx again matey ... :)
Haha, it isn't loaded, I promise! All for show-- as is the scowl!
@@TheoreticallyMedia 😝
Thanks for all your work! You are a great inspiration for me ;)
Every person ever talking about nodes :
Oh it's not that complex...
Then opens up a what looks like a map of the London underground with its 500 miles of rail network and 300 destinations 😭
Haha- I always call it spaghetti, but the london underground map is also pretty accurate! It looks complex on first glance, and then you think you've got it...only to realize: Wait, no-- it IS complex. haha.
On the plus side, when you build out a node system, at least it's YOUR mess.
I'm not going to call out the YT'er I'm referring to, as I do like them, but they always say the same type of thing about Comfy - it's not hard once you know it and so on. Yet, whenever they go live with it they spend half the time sorting problems and having things not work. That says it all.
@Elwaves2925 it's right throughout ... I was learning unreal and joined online course theyd say the same about materials using nodes.... Always "its not that hard when you look into it" then spent 1 hour going over it, skipping through parts by saying "we'll cover that in a later date"...
I call them "murder-boards".
@@achliscantplay4202 lol u mean like in a detectives office. The pegboard, the string, the suspects.
Flora looks exciting!
Do i need to buy kling/runway/luma credits/subscription to use the video models in flora/based labs or it's all included in the flora/basedlabs plan?
Gotta test these on my next video 😎
Totally-- I think they're great options that serve different purposes. The fact that you can hone in quickly on what model is best for a style in Based is pretty great-- and Flora might take you into some interesting story concept directions.
I just saw vidpanos test super pumped for that!!
This video popped off, no cap.
Bet. I'm vibin' on it.
In Flora, is that 15 minutes of just being on the site or 15 minutes of generation time or 15 minutes of video rendered out?
I also just saw in their price plan it mentions "hours of realtime". What is that?
15 min of ur gay
Is there any Ai right now that can add fx into my video that I have already? That's ideally what I'm looking for.
I have one on the list! It's a little janky right now, but you can see where it'll get to pretty quick: haoyuhsu.github.io/autovfx-website/
I tried to make consistent character in flora but couldnt in 15 mins... Anyone know is it possible?
So, I don’t think Flora is good at that yet. Best bet there might be to generate/train a model in Krea, Based, or really anywhere that model training is popping up. And then import into Flora as an image source.
I’ll reach out to them to suggest it!
yea we’re working on LORAs!
Thank you!
No cap indeed😂 0:58
BET!
New sitcom series on Paramount, ‘HAPPY TIM’ 😂
Hah, the gag of the show is everyone around me is happy, and I'm AI Grumpy!
@ 😂🤣😂
FLORA 🔥
It's a LOT of fun!
AI me would be cranky too if i had 7 fingers on one hand and 7.62 fingers on the other.
Couple of things:
1. How are all these platforms offering other models all-in-one? I don't get it. Are they tamed-down versions of Flux, SD, Runway, Kling, etc.? Something doesn't feel right.
2. Would appreciate slightly more in-depth one with Flora. 👍
Thanks as always, Tim. 💯
Flux, runway, Kling and others all provide their tools via a developer API, which sites like ours can use to add their tools. Hope that helps :D
Some of the commercial models are available to use via API, some of the opensource models could be deployed locally. With commercial models they usually just overcharge you (significantly) compared if you'd use the model on the original site - this is their profit. With open source models they charge for the used resources. Usually they not only overcharge you for resources but also use simplified versions of the models to generate faster and reduce resource usage. You'd never know what crippled version of Flux you use.
❤❤❤❤❤
Base labs looks good but a lot of apps that I will never need I looking to start doing ai generated videos I like Kling flux runway and their new talking app but I also want model training face swapping upscale etc clothing swap would be a bonus. Going this route I would have sign up for 3-subscriptions would there be a way to bring it down to 2 subscriptions? My budget would be around 40 per month. thanks love the videos!
Oh If this helps I use DaVinci resolve and premiere pro and I am pretty proficient at both.😂
I'm always thinking about switching to Davinci. I CAN use it-- I'm just really slow at it. Sadly, because of the nature of AI, and doing these videos, I never really have a chance to sit down and GET fast at Resolve.
Final Cut ALMOST got me interested in this latest version-- but, yeah...still gotta pass on that one.
(I'll say, the one killer feature for me in Premiere is still the Transcription Edit. For "talking head" videos like this-- I mean, it shaves HOURS out of my edit.)
@@TheoreticallyMedia lol but I still need to know which 2 sites would give me best bang for the buck to start ai video creation?
In your experience what 2 sites would be best?
Thanks
Ps…
I used premiere for for 15 years I now get it free from my sons edu subscription. But DaVinci is free or one time of 300. Forever!
If you can force yourself to do 3 projects in a row you will be more comfortable with it. It also has transcription editing like premiere I use it a lot too. Bonus it has smooth cut so if you cut in the middle of talking head you will not see the cut. Premiere has something like it but it’s a bit wonky sometimes. Ai in it now has auto voice isolation and music mix editor so you can turn up or down you can isolate every instrument take voice out of a song, just leave the drums take out the rest of the band sure no problem.
NightCafe is similar to Based and has been around since the beginning.
The problem I have with most platforms which offer a variety of models at your fingertips is the censorship. 🤬
At least with mystic ai access to Flux, there's minimal censorship 👍 👍 👍 Too bad mystic ai isn't a platform that offers the donut shop style selection of models that NightCafe offers. 😔
Nightcafe! Man, I haven't checked in on them in awhile! I'll swing over there this weekend!
@@TheoreticallyMedia
It's a donut shop where you can very conveniently select from a large variety of models. The NightCafe interface is excellent IMO
@TheoreticallyMedia
NightCafe is still quite relevant in the AI space....... considering how old they are, this is quite a feat!
They thrive in the community aspect which for those who need or love that, is great.
I, as a secret misanthrope, am not enamored by that part of NightCafe
😉
@@LouisGedo NC, like donuts, is very addictive! 😋
@diane161061
And who doesn't like donut shops where you have a nice variety of different styles to easily choose from right at your fingertips!
That's why I do like NightCafe....... it's like a donut shop.
These workflow GUI are kind of dumb in my opinion, They would easily be able to change this into a combo boxes with options or something similar for ease of end user and would appeal to a lot larger audience.
For me, the elephant in the room is lip sync. You either get good lip sync which lasts for longer than ten seconds with very little character motion, or good character motion with little to no lip syncing (other than maybe ten seconds or less). I'm waiting for that (Runways Act one is close).
I've got a look at Lipdub.AI on the docket. That's the best one I've seen so far.
Make a GPT for your dad jokes Tim 🥺
Haha, I'll admit, every once in awhile I'll ask ChatGPT for a few jokes, and-- not gonna lie, it's got some zingers for sure!
Flora's capabilities look fun, but the murder-board does not 😅 I hope this stupid trend dies off, or, at least, they start providing more neurodivirgent-friendly organisational options (folders, trees, etc.)
I already spend around $400/month on this hobby, so it is going to take more than another aggregator to get me to switch, or even try new stuff.
Haha, the biggest thumbs up and heart I can possibly give for "Murder Board!"
@TheoreticallyMedia Thanks, Tim! Means a lot, coming from the Dad of AI Dad Jokes 🤘❤️🙇♀️🇦🇺🐨🦘 (I am also kinds famous amongst my students for mum jokes - both of them).
Hey Weber here I’m the founder of FLORA - we’re working on more neurodivergent friendly organization options. We hope to add everything in one place in a new type of workflow-oriented UX so in addition to acccessing everything, you can stitch them together and automatically run a chain of FLUX > Image to Video nodes with different prompts at the same time, and share your projects with other people with real-time multiplayer. We’re just in Alpha so hopefully we justify over time our node-based approach :)
I find that none of these video models do black faces all that well. Well they do for some, just not mine. They always change my features... drastically
Yeah, they always turn me asian. It obviously morphs toward the training data at some point. But hey, maybe if the actor is super handsome, that's ok?
@😂😂😂😂😂😂😂😂
@ it does kinda ruin it for me when I want to cut in between a live actor and an AI one for cgi type scenes
Nice video!
But most of all-in-one platforms are so ridiculously overpriced compared to the pricings of the original tools they provide functionality of that anyone who managed to graduate the elementary school won't touch them with a 10-feet pole.
And carry over credits made me chuckle. Something like "Here are 1000 credits for only 99.99, one generation on Flux cost 10 credits and one Kling generation cost 50 credits but you can carry over month". Yeah, good luck with that.
Eventually there will be a google one, that already has all my images sorted by faces, so it should be able to perfectly recreate me, friends and family and all the places I often photograph.
Then I can say "ok google you know were we've been the last 10 days (city tour Paris) please make some photos and videos from me and my girlfriend that fit the places we've been to (time and weather correct) please."
They actually have Vue already (I think that’s what it was called) but, typical of Google, they’ve forgotten all about it.
Please put more effort in your clickbaiting. Your video title sounds very truthful and informative. I suggest you to append "I Quit RUclips". 😅🤣😂
Haha, damned if you do...damned if you don't! Ok, "These Two New AI Platforms Just KILLED Midjourney!"
👋 hi
Heya Louis! Curious to your thoughts on Flora! It's pretty neat and for sure leans into some of the wacky creative side of AI!
Ai Santa 😅
and it isn't even December! Speaking of Santa-- keep an eye out for today's video!
@@TheoreticallyMedia You have a huge research On Generative AI . Learned a lot from you . 👍👍
Hawk Torah
Tony Hawk's Pro skater
@@GalaxVerseSkate on that thang ya get me?
@@alexfilma16 nah ya don't get you
@@GalaxVerse Ok.
Two tools I Should know about??? Really???
14 free credits? $25 for 1,000 credits with a basic result. Seriously???
I hope they paid you a lot of very good money'
Because you're probably familiar with the saying: Tell a lie once, and all your truths become questionable.
I know this sounds a bit dramatic, but I feel betrayed. :(
There is way too much information about too many things in this video. The rapid pace that you go through is not helping the average person.
there are literally 2 things in this video.
It's in the title.
Watch the video in 0.5x speed pea brain
Ur just gay is all
@@mainaccount888 :Oh, and a child chimes in !
I'm early😭✌🔥
Ok....this is an achievement for you?
China is so far ahead
Rumor is that Sora releases sometime this month. Literally just asked David from Midjourney if Video was shipping before the end of the year-- he said probably closer to Jan, only because the team doesn't want to work over the holidays (totally understand that).
Sora actually exists 😅 nice
It looks exactly like you. Get over yourself. (quite good looking actually).
Haha, I appreciate that! Hmmm, maybe that's why AI me is always so cranky? He's like: "Get over it! I look like you!"
Gay