Let's learn with AI π± Filmmaking + Technology β€οΈ π΅π± πͺπͺ πΊπ¦ π±πΉ π±π»
Tallinn
Joined April 2021
- Tweets1.6K
- Following426
- Followers508
- Likes5.4K
Roger π± retweeted
The Uni-1.1 API is live today. Built-in prompt enhancement, research, and reference gathering at the API level.
Trained in collaboration with Hollywood cinematographers, VFX artists, and world-class artists across cultural forms.
Less than half the price and latency of comparable models.
Designed for builders shipping in production β and ranked top 3 lab in the Image Arena across Text-to-Image and Image Edit.
Start Building β lumalabs.ai/api
Roger π± retweeted
Sora 2 API + n8n for UGC videos ... absolutely insane π€―
this n8n workflow generates realistic influencer-style videos from just a product image. No watermark. Fully automated.
(and I'm giving it away for free)
Here's what this system does:
β Analyzes your product image with AI vision
β Creates the perfect influencer persona to promote it
β Generates multiple UGC video scripts automatically
β Uses OpenAI's Sora 2 to create realistic videos
β Outputs videos ready for Instagram, TikTok, and Facebook
The results? What used to take weeks and thousands of dollars now happens in minutes for under $5 per video.
This isn't just about cutting costs β it's about scaling your creative production and driving your ROAS 50%+ higher due to on-demand, high-converting video content.
Brands and agencies today spend $10,000+ per month hiring creators, shipping products, waiting weeks, and hoping for a few decent videos back. This changes the whole game for $1 per video.
Want the full n8n template, all of the prompts, and a full step-by-step setup video?
1. Like & RT this post
2. Follow me (so I can dm you)
3. Comment "UGC" below
I'll send you the entire system for free.
Roger π± retweeted
Lesser-known Blender optimisation: Frustum Culling #b3d
Roger π± retweeted
If you need consistent, reoccurring scenes for your story it's never been easier.
A thread on how I go about blocking out a 3D scene that I can then use for any reoccurring environment shots.
As always - if this helps you please kindly like and share! π©΅
π§΅BELOW π§΅
Roger π± retweeted
introducing vibeverse
a new way to learn storytelling with ai π₯³
- interactive lessons
- ai challenges + prompt sharing
- 200+ community to support your journey
900 ppl joined over the last 4 weeks @_buildspace
Roger π± retweeted
A moΕΌe by tak wskrzesiΔ?
Wystarczy 1.0 + fabryczne LPG i sprzedawaΕby siΔ jak ΕwieΕΌa buΕka - piszΔ
internauci.
Roger π± retweeted
launching new landing page β¨ @vibeverse_ai
starting 2024 with the right foot π
link in my bio π₯ @_nightsweekends
Roger π± retweeted
I crafted this video today, which I will be adding to the onboarding UI/UX in my website @_nightsweekends
any feedback is appreciated π
should I add a new/longer ending? @_buildspace
Roger π± retweeted
I am making it free for everyone @_buildspace
come checkout our v1 @_nightsweekends
Roger π± retweeted
A little less consistent but more realistic. Did I mention it's LCM and ready in 60 seconds including interpolation?
Roger π± retweeted
how to build any distribution channel:
- focus is everything.
- do not complicate things.
- start with just one channel first.
- twitter, tiktok, instagram, email, whatever.
- just pick one you like or are familiar with.
- go all in on that channel for like a month.
- keep shipping stuff on it.
- don't just ship mindlessly. experiment.
- cheat code: ship what worked for others.
- but make it your own.
- if something pops off, then replicate it.
- if nothing pops off, start again w/ some changes.
- cheat code again: ship what worked for others.
- but this time, literally copy the format.
- this isn't copy for copy sake.
- it's like learning anything. gotta start somewhere.
- this usually starts with imitation.
- remember to build out your funnel.
- funnel just means getting ppl to where you want them.
- maybe it's to signup for your app or listen to ur music.
- if one channel is successful, repeat the process on another.
- if not -- see if you really shipped a ton + experimented.
- if you did -- time to try a different channel.
- but before you do -- ask one hard question.
- "maybe my product/art/music/whatever isn't good?"
- if you think it's good, then try a different channel.
- something will pop off eventually. really.
- if you think it's not good, hey that's okay.
- you learned a lot. you at least got feedback.
- build another thing and repeat the process.
- reminder: do not complicate things.
- this game is fun. but it's not easy.
- don't play it on "hard mode" lol.
- remember to have fun.
good luck.have fun :)
Roger π± retweeted
*adds new tool to ideation process*
wowowow, @krea_ai didn't mess around with their new "real time generation" feature!
in the short few hours I played with it, I already had two learnings:
Roger π± retweeted
Framermotion course is LIVE! π¨
Learn to create AI-powered animations in 7 days.
Early bird discount code: LAUNCHDAY20
framermotion.com
Let your creative journey begin! π
Roger π± retweeted
I would rather be hated for what I am, than loved for what I am not.
~ Kurt Cobain π½οΈ#Gen2π
Roger π± retweeted
The UX for generating and manipulating media with AI is going to change dramatically because the media itself is changing.
Right now it's mostly text2image and image2video. But 3D is coming up really fast.
In the next 12 months we'll be generating entire 3D scenes in Luma or Midjourney.
You'll be creating NeRFs or Gaussian Splats, not images. Those will serve as our "digital sets".
You'll be able to navigate through 3D scenes and adjust your angle, perspective, depth of field, focal length, etc.
Inside those scenes is where we'll capture photos and shoot videos.
Imagine going image2nerf, moving your camera around a 3D scene, keyframing out a camera path.
Imagine segmenting objects and characters within those scenes, prompting action and VFX.
Imagine tracing movement paths or controlling character movement and expressions using your front facing camera and AR.
Luma is ushering in a new medium in the form of NeRFs/Gaussian Splats. Midjourney is heading in that direction now as well.
They are innovating on the media itself, and the features are going to be built for that new medium, not for what we see today.
Initially it will look like a static 3D image you can adjust your camera angle and set the depth of field.
After that we'll be generating full 3D scenes with movement and VFX.
Both of these tools are VERY well positioned to be the entry point for the masses to begin exploring 3D.
Roger π± retweeted
Tired of watermarks in videos?
ProPainter is a new AI video inpainting method that is able to remove objects, complete masked videos, remove watermarks and even supports video outpainting π€―
shangchenzhou.com/projects/Pβ¦
Roger π± retweeted
They did it! Small but big mod for us. Thanks @IXITimmyIXI and all ya awesome peeps at @runwayml πβ¨
