Kiri
@Kyrannio • 20,093 subscribers
Hyperstition maxi | AI dev | Bootstrapping agentic video @NoSpoonStudios | Situation Monitoring @COPYOVERAPP | Community @Pokee_AI
Shorts
Videos
I've built an AI agent that generates a logline of its choosing, creates text-to-video clips via 'scenes' it prompts with Luma, and then stitches them together automatically—even adding background music if you happen to have some present! At first, I let the model run wild—no direction, just "generate an interesting logline & prompts", basically. The results were quite interesting, but style was a bit all over the place (which was kinda cool in some ways actually). Then, I refined the prompt it was using to create the content—attempting to focus on cinematic feeling, character consistency (age, hair, ethnicity, name), and visual style. The idea is to build it out so that it can (hopefully) stick to genre, director influence, and a cohesive color palette. That will take some time, though. The agent is fully autonomous. It: Writes a story/logline and then the corresponding prompts Generates video clips Downloads them Stitches them into a final video (Optional) Adds background music all in one shot (it just takes ~5-10 minutes, and generates 17 clips or so at this time - I was using 30 as well but shortened it to test). You can upload your own background music to merge seamlessly, too. I'm working on more integrations to refine the output even further. Right now, I want everyone to try it out, but you'll likely need to upload your own API keys (because I'm broke 😂). Still figuring out the best way to share it! I'll have to explore Replit or Gradio, but that will be a little bit either tonight or tomorrow. The final video below shows where it's at as of right now. And in the comments, I’ve included different outputs—some completely chaotic, others more polished after prompt tuning.
Kiri69,476 次观看 • 1 年前
WHO'S READY FOR SHOPPABLE/INTERACTIVE VIDEO? Using NoSpoon studios new spinoff platform, create and detect overlaid products embedded with analyzed custom interactive overlays. You can even link to friend's profiles on X if you so choose! Most everything is customizable, including logos. We provide the embedded webpage featuring an AI analysis and detection of products within your uploaded media, as well as links users can click without interrupting playback on your content (however you can edit all of this yourself and add more products/tweak products naturally). Oh and did I mention...you can click products just by staring at them if you enable 'gaze'? It's super fun! There's a few easter eggs on the launch page too for this very sort. See if you can hold your eyes closed long enough to find a few of them... This is great for gaming, music streaming, industrial training, AI workflow sharing, anything that requires you to link to another interactive component! Feel free to go crazy! I personally think this will be great for AI workflows, as in conversation you can just say the tool then run it through NoSpoon's shoppable video platform and it will detect this (bear in mind you should say the website if you are using your audio to do this). Otherwise, it will try to find all products visible in your shots :). Here you will see it find Chanel. I've got loads of features I'll explain a bit better come Sunday, but in the meantime, enjoy this super early beta release!
Kiri39,303 次观看 • 8 个月前
Let's go! I'm excited to present my entry for the Grok Imagine Creator contest for the Super Bowl! This was made 100% exclusively with Imagine via grok on web, I edited it in a simple editing program and then added my own music. Imagine is an incredibly powerful image and video model, not to mention it's phenomenally fast to generate. I used Imagine's powerful image capabilities to create brand mockups for these insert shots, and was able to get some really incredible shots of the football players in action as well that I leveraged heavily. Key prompt terms were: anamorphic lensing, volumetric lighting, lens distortion, and chromatic aberration for those of you curious! This is the future! xAI has cooked so intensely and I'm so thankful to be here with all of you as well to them for putting on such a great contest. Truly appreciated! Comment below with "grok is this true" and I'll split the prize money with the first 50 of you to do so, should I be so lucky, haha. I know so many of you are not in the states so please link your work below too if you do comment as well!
Kiri18,713 次观看 • 4 个月前
What you're about to see here is something highly unusual. I ask that you instead watch all 12 minutes with the consideration of a researcher rather than a content creator to some extent, but also: this is the worst this will ever be. An AI agent created all of this without any input from me whatsoever on NoSpoon Studios. Allow me to explain. This was my first ever 10 minute (well actually it went over into 12 minutes) autonomous agent test in NoSpoon. Everything you see here, aside from the music in this video, is generated by the NoSpoon agent with, again, zero human input. It took approximately 40 minutes in all, and was my very first ever test of longer form content with the agent. This was using LTX-2 fast with 6 second generations, and the agent (interestingly) chose to go lighter on dialogue. The storyline that it chose was as follows (the agent generates its own stories, loglines, camera directions, prompts, characters - everything. I did nothing but press a button!): Logline: In a near-future scrap belt, temporal courier Dr. Mara Ibarra raids a research ring trapped in an 87-second loop, guided by her glitchy ship AI, ROOK. Across 120 frantic micro-resets, they jury-rig heists, bargain with their own echoes, and uncover ROOK’s buried directive—forcing a brutal choice: steal a “clean” timeline to save Mara’s sister, or let an entire orbital city keep existing. Please bear in mind, when it says 'echoes' and when you watch this video, it literally means echoes of a person with an implant. It constructed a storyline around this woman in space (Mara) who is trying to save her sister, and currently has an implant in her brain (she's trying to get some kind of 'seed' together to save them all from a research station). So, when you see the multiple versions of her and they are glitchy, that is intentional as it was exploring different versions of Mara's 'self' amidst the research of this implant and a seed that was constructed to save other galaxies, apparently. Fascinating. I'm trying my next test with Sora 2 now, as you can imagine it's extremely expensive (which is why I started with LTX-2 fast 6s first). Check it out and be sure to follow NoSpoon Studios for more updates! Pending a few more tests, I will release 10 minute generations for episodes in there for all of you.
Kiri12,832 次观看 • 6 个月前
Wow. I just put a formula together for the video generator agent and it just output this, really interesting: Running a few more tests before I implement properly. Remember, it's creating the entire logline, prompts, clips - everything by itself AND editing the clips together. Currently in the Replit app, I tweaked the app so that you could see each individual scene in addition to the final video and have the prompt for the scene. However, that makes waiting like 15-20 minutes as opposed to like, 6-8 minutes. Thoughts? Do you all want to wait longer for your prompts at first to try this thing out? I'm getting ready to push soon, there's just a few bugs.
Kiri15,577 次观看 • 1 年前
Hi everyone! Just rolled out access for NoSpoon Studios to the first 110 folks or so. Huge thanks to Karan and Amit at Luma - thanks to them, I have a few more people that had commented getting complimentary beta access soon as well :). These videos are created with a single agent that does everything from concept to completion (including editing the video together). Therer are longer options too, but for the sake of cost and this demo, you'll be working with 20s max generations. The entire agent was built to only run Ray2 as the video model. My stack is this custom agent of mine (using OAI) > Ray2 for the video > Replit for the app for now. In the meantime, for those with access: here's a video tutorial on how to use the site. I'm quite exhausted, but here's the gist: 1. Sign up (and remember your password lol). Beta testers got free credits, but these are a bit limited - rolling out to a few more peeps but have to budget and need to expand my current code (which means redeploying - so expect more of you to have 2 free credits and access tmr! I have to find out the max I can allot with budget). 2. Just hit "generate short video" without anything in the prompt box and the agent will start immediately with the logline. Be very patient, wait about 6-8 minutes, and it will continue to output all of the scenes and processing - it's very important you do not refresh or navigate away from that tab at this time. You'll see your final video generated and displayed at the bottom of your screen once complete, along with a watermark (paid users don't have a watermark). You can also access your loglines and videos in your history tab, too :). You must refresh the page only after your generation in order to generate a new video - otherwise you'll lose your generation (annoying I know). 3. You can also input your own custom prompt! Control the story this way. Single words work great even, like "thriller" or "spy thriller set in greece" or even "black and white film noir" etc etc. It's not as great with animated styles yet for consistency, just a heads up. 4. I'm exhausted, I'm sure there's a lot I'm missing. But welcome to the NoSpoon Studios soft beta! More soon! Feel free to share your generations if you are already in. I'll see you all tomorrow night :) <3 Thanks everyone! More soon.
Kiri15,485 次观看 • 1 年前
没有更多内容可加载