AKT

518 posts

AKT banner
AKT

AKT

@akirathompson

Group Creative Director @ Publicis Production. Ex-R/GA CD. Ex-Framestore Design Director. AI / ML fanatic. @aktions on IG

Katılım Mart 2014
352 Takip Edilen137 Takipçiler
AKT
AKT@akirathompson·
Exploring vibe-coded motion graphics. Everything below was built with @cursor_ai , @claudeai , and @Remotion , no After Effects. I describe the animation, Claude codes it, we iterate via notes + revisions. Since it’s all code, changes are precise and everything else stays intact
English
3
0
3
428
Lovis Odin
Lovis Odin@OdinLovis·
After long months , I am so happy to share probably one of the biggest and most early project using mostly ComfyUI (in march) and other AI tool to generate a full animation movie. this project was a Real team project made with @Dogstudio for @CocaCola . Second part IS making off ! Hope you will like it !
English
26
71
467
33.6K
AKT
AKT@akirathompson·
@cocktailpeanut This is super cool. Tried installing on Pinokio but getting 'Error, connection errored out'. Any ideas what the probelm is?
English
0
0
0
104
cocktail peanut
cocktail peanut@cocktailpeanut·
Your machine may not support the bfloat16 device type, in which case the app would have just hung forever, taking up all the VRAM without doing anything. I've pushed a fix for this case so update and retry if this was the case x.com/cocktailpeanut…
cocktail peanut@cocktailpeanut

By the way, some old machines don't support bfloat16, which caused the inference to hang forever. A radio button has been added to let you select between bfloat16 & float16. Update and retry if it didn't work before. The default is bfloat16, but if it doesn't work, try float16.

English
2
0
5
1.3K
cocktail peanut
cocktail peanut@cocktailpeanut·
Run CogVideoX Locally [NVIDIA ONLY] CogVideoX lets you generate 6 second videos with as low as 5GB VRAM. Releasing a 1 click launcher for the official gradio app from the CogVideoX project.
Gradio@Gradio

📢🔥Hot New Release: CogVideoX-5B, a new text-to-video model from @thukeg group (the group behind GLM LLM series) - More examples from the 5B model in this thread👇 - GPU vram requirement on Diffusers: 20.7GB for BF16 and 11.4GB for INT8 - Inference for 50 steps on BF16: 90s on H100 and 180s on A100 - open weights - Video length: 6s, Resolution: 720 x 480 - Best news: Previously released CogVideoX-2B is now switched to Apache 2.0 from previous custom license! 🥳🎉🍾

English
11
42
237
81.1K
AKT
AKT@akirathompson·
@aishamadeit Would have been better if it was done in cg?
English
0
0
0
61
AKT
AKT@akirathompson·
@aishamadeit Controversial hot take: it’s the human creative/art direction and comp work that makes it bad, not the AI. 😬 fight me 😅
English
0
0
0
104
AKT
AKT@akirathompson·
@eyedesyn Controversial hot take: it’s the creative/art direction and comp work that makes it bad, not the AI outputs. 😬 fight me 😅
English
0
0
1
137
AKT
AKT@akirathompson·
@emmanuel_2m Looks very similar to a CSM or MeshyAI for 3D model > animate in Mixamo or Anything World workflow. Still nice to have it all contained on one platorm if it performs at or better than the aforementioned
English
0
0
2
211
Emm | scenario.com
Emm | scenario.com@emmanuel_2m·
Well, well. AI-generated character turnarounds might be more useful than I thought. 🧐🤔😁 Working on something new. #Scenario
Emm | scenario.com tweet media
English
27
46
339
42.3K
AKT
AKT@akirathompson·
@PurzBeats 100% MTV would have run this during Liquid Television back in the day
English
0
0
1
68
Purz.ai
Purz.ai@PurzBeats·
😶‍🌫️
GIF
QME
30
49
447
8.5K
AKT
AKT@akirathompson·
@PurzBeats Any more information on how this is made 🙏🏻?
English
1
0
2
36
Purz.ai
Purz.ai@PurzBeats·
🤔
GIF
QME
5
7
67
1.8K
AKT
AKT@akirathompson·
@Aaron_porter @Oddernod @fable_motion @rive_app @Cascadeur_soft Sounds like a job for chatgpt 😄. As far as styles, wooly/embroidered, cut wood/jigsaw and 8-bit video game would be kinda cool. Or even an option to upload an image style reference. Love the tool so far, nice work
English
1
0
1
98
Ryan Summers
Ryan Summers@Oddernod·
{ motion design } I go away for a couple days and: • PROJECT AVALANCHE drops from Unreal • @fable_motion Prism is freaking everyone out • @rive_app is GDC blowing minds at GDC • @Cascadeur_soft new features 🤯 • new Clip Studio Pro • Dope new 3D features and UI speedups in After Effects • Controversial AI stuff from Substance It’s all moving so dang fast 🏎️💨
English
8
7
98
9K
Aaron Porter
Aaron Porter@Aaron_porter·
@Oddernod @fable_motion @rive_app @Cascadeur_soft Ah, gotcha! Our team shares your cautious sentiment fwiw... Prism is pretty darn magical at moments, but still discovering how it fits into real-world workflows. Very much an active project, so we're all ears. Curious your thoughts when you get the chance to dive in!
English
2
0
0
163
Cartoon Brew
Cartoon Brew@cartoonbrew·
Protip: In animation, voice acting pays better than making the artwork.
English
18
11
141
69.3K
Chris Vranos
Chris Vranos@ChrisVranos·
A pigeon just very intentionally shit on me so I'm declaring war on all pigeons.
English
2
0
4
428
AKT
AKT@akirathompson·
@flytip I bought 2 Popeye’s meals over the weekend. $33. Blew my damn mind. Still tasted delicious.
English
0
0
0
73
Richard Ting
Richard Ting@flytip·
I found out today that a Big Mac meal costs $18 and nothing makes sense to me anymore.
Brooklyn, NY 🇺🇸 English
2
0
6
1.2K
AKT
AKT@akirathompson·
@TrailersFromHel @joe_dante All time favorite. Might have to run it back tonight. Hey Pinocchio! Where you going!?
English
0
0
1
24
AKT
AKT@akirathompson·
@LinusEkenstam Is it just me or do these robots look janny af? All the little grid square bots barely move without crashing into eachother, the big arm bots move half as fast as a human would and don’t get me started on the humanoid bots, they look like Boston Dynamics from 15 years ago.
English
0
0
2
209
Linus ✦ Ekenstam
Linus ✦ Ekenstam@LinusEkenstam·
Amazon got more than 750.000 robots deployed. Most people don’t realize how fast the robotics industry is scaling Amazon is the perfect candidate. 10 years ago, robots were practically non-existent in their global warehouse and distribution network. But this is the actual acceleration ramp-up. 2013: 1,000 2014: 15,000 2017: 100,000 2019: 200,000 2021: 350,000 2022: 520,000 2023: 750,000 Let’s zoom in on the two last jumps. 400,000 additional robotic units in roughly two years. That results in thousand of new units deployed *every week* 🚀 It’s clear. Beyond doubt. That AI, robotics, computer vision, will and is replacing a lot of human labor. And will continue accelerating that progress in the next decade. It’s important to note that this will also accelerate the need for more high-skilled work. Make industry safer. The biggest challenge we will face is the grandeur of re-skilling and up-skilling that will be facing the workforce in a relative short period of time ahead. We’ve entered the Era of Robotics and The Age of Intelligence all at once.
English
187
977
4.1K
1.2M
AKT
AKT@akirathompson·
@ChrisVranos @Arcane_Motion Wonder if you could figure out how to upscale the nightshade effect on the original so that when it’s downscaled to 512 and compressed to jpg or webp that is when the nightshade is correct over the image
English
1
0
0
70
Chris Vranos
Chris Vranos@ChrisVranos·
@Arcane_Motion You could definitely poison a data set if all your work was released with lossless compression and at the right training scale (512x512). But most websites compress to jpg or webp right away.
English
1
0
0
109
AKT
AKT@akirathompson·
@PurzBeats As an advertising Creative Director specializing in motion graphics, 3D and now gen AI, I have this conversation on a daily basis
English
0
0
1
18
Purz.ai
Purz.ai@PurzBeats·
I have had SO MANY meetings with "creatives" looking to add AI to their workflows (A&R Reps mostly) and after 30 min in the call when they find out it's not just typing words into a website and having fully bespoke music videos pop out, they suddenly need to 'get back to me.' 🤣
English
36
14
203
11K
AKT retweetledi
Runway
Runway@runwayml·
Today we're thrilled to announce that we are partnering with @GettyImages to build enterprise ready AI tools that will allow companies to create high-quality, customized video content. Read more: runwayml.com/blog/runway-pa…
English
64
127
708
196.2K
Bilawal Sidhu
Bilawal Sidhu@bilawalsidhu·
This is image-to-video with Stable Video Diffusion (SVD). The temporal coherence is impressive! The model is also well suited for multi-view consistency (i.e. doing image-to-3D) Look fwd to diving deeper and sharing more. Meanwhile check out @fofrAI
English
15
21
170
20K