GPU Whisperer :: instagram.com/_hackmans_ R&D for @PrefixStudios
Joined July 2019
- Tweets 1,029
- Following 540
- Followers 1,642
- Likes 9,012
134 Photos and videos
Pinned Tweet
19 Feb 2023
New video of my game @talesofsyn showing GPT3 based NPC chat interaction with pre-prompting and custom voice synth by @elevenlabs and #vfxgraph weather system background shaders made with AI generated wet maps
#madewithunity #gamedev #indiedev #indiegame #unity3d @GeppettoAi
10
21
100
14,580
-=HACKMANS=- retweeted
May 26
FLUX.2's @bfl_ml text tokens aren't just holding your prompt.
During image editing, they absorb reference image content, and some of that absorbed content, like color and style, causally drives the output appearance.
New paper 🧵👇
7
35
203
26,161
-=HACKMANS=- retweeted
Apr 16
493
4,886
22,127
1,591,232
-=HACKMANS=- retweeted
Apr 9
Dialing in real time V2V in Sentinel.
No Python or venvs, just an installer. Runs locally on any 16GB NVIDIA GPU.
Sign up for early access: ood-labs.com/sentinel/waitli…
42
89
815
57,565
-=HACKMANS=- retweeted
Feb 27
Autechre Guitar is out today!
shaneparish.bandcamp.com/alb…
This has been a truly magical time. From the making of this record, to seeing how many people it has connected with, and now as it leaves the nest. I’ve never experienced anything like it. 🙏
15
72
495
64,529
-=HACKMANS=- retweeted
25 Nov 2025
FLUX.2 is here - our most capable image generation & editing model to date.
Multi-reference. 4MP. Production-ready. Open weights.
Into the new.
162
470
2,953
482,543
-=HACKMANS=- retweeted
21 Nov 2025
Ok, folks. Here's Voxel Duke 3D Version 1.0!
I'm gonna take a break from the internet for a bit. Have fun, I hope everything works.
moddb.com/mods/voxel-duke-nu… via @moddb
47
356
2,001
106,235
-=HACKMANS=- retweeted
13 Aug 2025
Real-time visual effects, running 100% locally in your browser! 🚀
I created a demo where you can just drop your video and watch it analyze feature points and optical flow for some cool visual effects!
Have fun 😊
Demo: whenistheweekend.com/vfx.htm…
#ComputerVision #VFX
12
75
570
33,504
-=HACKMANS=- retweeted
7 Aug 2025
BeatBandit - New era for storytellers
Turn a raw story idea into beats, scenes, treatment and a screenplay draft - 100 × faster, automatically applying professional screenwriter techniques, without losing control of your voice.
It’s happening!
🧵Read all about BeatBandit
23
32
321
59,808
30 Jul 2025
Excited to share that Showrunner is now live!! I’ve been helping to build the platform alongside an incredible team at Fable and can’t wait to see the community start shaping this new entertainment medium!
#showrunner #netflixofai #aistorytelling
30 Jul 2025
Introducing Showrunner: the Netflix of AI
From our South Park AI experiment to today we’ve believed AI movies/shows are a playable medium.
We just raised a round from Amazon & more and the Alpha is live today
Comment for an access code to make with all our shows.
2
3
12
870
28 Jul 2025
More AI VFX experiments with Wan2.1 VACE. Art directable explosions on existing footage plate. Being able to shape the effect and get all the free secondary motion of the environment interactions is phenomenal
1
4
223
26 Jul 2025
Some AI VFX experiments to learn Wan2.1 VACE inpainting. We really are spoiled with the level of creative control possible from open source AI video
2
1
10
355
-=HACKMANS=- retweeted
10 Jul 2025
IlluGen will be out on July 28th! Set your calendars for Elemental Week 2025 (July 28 - 30th). We'll be streaming the latest news on our Youtube Channel!
10 Jul 2025
For years I've wanted to be able to make looping, tiling smoke textures from EmberGen. IlluGen finally allows me to do that without even having to think about how to do it.
I made a square, looping smoke simulation flipbook in EmberGen. I imported it into IlluGen as a flipbook, ran a sheet decode on it to give me individual frames, used a Voronoi tile node and re-encoded the sheet for an absolutely perfect tiling, looping simulation.
1
22
169
13,124
12 Jun 2025
Tested #MultiTalk AI video (wan2.1) generation driven by audio, using a still image of Kurt and audio clip from Unplugged. Uncanny, but pretty impressed with the emotion and performance. 6 * 81 frame chunks, 47GB VRAM, 17min per chunk on RTX Pro 6000
4
4
12
1,576
-=HACKMANS=- retweeted
24 Feb 2025
What if an AI agent makes a phone call, then realizes the other person is also an AI agent?
At the ElevenLabs London Hackathon, Boris Starkov and Anton Pidkuiko introduced a custom protocol that AI agents can switch into for error-proof communication that's 80% more efficient
It's mind-blowing
70
183
791
132,730
-=HACKMANS=- retweeted
6 Feb 2025
Apple gets it. Robots are going to be everywhere, but they won’t look like robots. Check out their new paper ELEGNT.
I believe this is the future of everyday objects: helpful and human.
224
803
7,656
1,031,398
-=HACKMANS=- retweeted
3 Feb 2025
🔥Introducing #𝐌𝐚𝐭𝐀𝐧𝐲𝐨𝐧𝐞 for human video matting!🔥
🤡 Fast video matting with customizable target
🤡 Stable human tracking
🤡 Fine-grained matting details
- Paper: arxiv.org/abs/2501.14677
- Project: pq-yang.github.io/projects/M…
- Code (soon): github.com/pq-yang/MatAnyone
17
96
636
42,507