AI Creator | Exploring the future with ChatGPT, Seedance 2.5 & CapCut. Creating innovative AI content and digital experiences

USA
From product concept to promotional content, the Tesla Roadster is a great way to explore a different approach to video creation. Describe the style, message, and overall direction you have in mind, then let the CapCut Web × Codex workflow help carry it forward. Natural language becomes the bridge between your idea and the editing process: Codex interprets the brief, and CapCut Web helps execute the video creation. A more connected way to turn creative intent into content. #capcutweb #capcutchatgptplugin #capcutcodex #capcutmarketing
4
3
4
188
for robot training you need loads of different 3D rooms with people in them. Viggle's JST-2 makes one in seconds for under a cent of compute
Today we are releasing JST-2, the world's first full-stack physics-native world foundation model. Many world models learn in pixel/video space. We went back to first principles and built the entire model stack in 3D. Generate a 3D scene in seconds and expand it in real time. Create characters with motion built in, all through one general architecture.
1
56
59
40,259
CapCut Black Friday brings 60% OFF Pro and 40% OFF Ultra! Oct 20–30, only on CapCut Desktop. Who says a Black Friday ad has to be a boring sales pitch? I used CapCut PC to turn a creative idea into an eye-catching promotional video, then polished it with multi-track editing, effects, and precise timeline adjustments. From AI-generated visuals to a finished campaign, the entire creative process comes together in CapCut PC. #capcutpc #capcutmarketing #capcutsale #capcutblackfriday
54
46
54
37,018
Long conversations shouldn’t require paying the full cost for the same context repeatedly. With Inworld’s caching, repeated tokens on Gemma 4 26B are priced at just $0.035 per million, down from $0.07. That’s a 50% cost reduction, making large-scale consumer AI more affordable.
Inworld is your new favorite inference provider. We already process tens of trillions of tokens a month because we understand consumer AI different, and we've built our inference from the ground up for it. Today, the inference top consumer AI apps run on is opening up.
1
38
41
35,167
Up to 60% OFF CapCut Desktop, starting Oct 20. Desktop only. Sometimes the smallest distraction can take attention away from an otherwise great marketing shot. Maybe there's an unwanted object in the background of your AI-generated footage, or something in the frame doesn't fit the look you're going for. CapCut's Object Removal helps clean up distracting elements and make a shot feel more polished. For a trend-inspired video like a Jean Phil character campaign, cleaner visuals can help keep the audience focused on the character and the creative idea. And because you can continue editing in CapCut, you don't have to treat every generated clip as a finished product. There's still room to refine the footage and make it fit the larger video. #capcutpc #capcutmarketing #capcutsale #capcutblackfriday
101
211
220
68,443
Same dragon, every shot. You'd think the sound would agree, but in the stitched version it shifted with every cut. Mirelo SFX 2.0 keeps the dragon sounding like one dragon from cut to cut, which is the part separately generated shots rarely manage on their own.
AI Videos finally sound real. Introducing Mirelo SFX 2.0. It watches your video, cuts the music you never prompted, keeps the dialogue, and builds a full sound design synced to the action and consistent across any length clips. You’ve made it look right. We’ll make it sound right. Live today through our API, on @runwayml and @fal. Other platforms coming soon. (Thread)
1
38
43
26,754
MR. ANDERSION Official retweeted
AI Videos finally sound real. Introducing Mirelo SFX 2.0. It watches your video, cuts the music you never prompted, keeps the dialogue, and builds a full sound design synced to the action and consistent across any length clips. You’ve made it look right. We’ll make it sound right. Live today through our API, on @runwayml and @fal. Other platforms coming soon. (Thread)
452
412
3,835
1,880,627
MR. ANDERSION Official retweeted
Your next edit just got cheaper. Save 60% on Pro or 40% on Ultra, Oct 20–30. CapCut Desktop only. I wanted to test: How Codex works in your editor. Inside CapCut PC: Codex creates the AI character. AI Video brings it to life. Automated editing builds the first draft. Then creators take over with final edits. #capcutpc #capcutchatgptplugin #capcutcodex #capcutmarketing #capcutsale
143
212
241
72,828
What does the learner actually get to decide? Everything that matters. The coach sets the frame, but the learner supplies the unexpected details.
What should humans learn in the age of AI—and how should we learn with it? Today, we’re officially launching Gakku—an AI-native learning coach for writing, speaking, and thinking, built for ages 6–16. As natural language becomes our interface with intelligence, critical thinking and original expression matter more—not less. Most AI products optimize for results. Gakku is built around the process—the human experience of getting there. Through real-time voice and multimodal interaction, learners talk it out, explore ideas, revise, and create something of their own. Meet Gakku: gakku.ai/
88
79
86
31,814
Product education is getting smarter. @YouArtStudio Agent helps turn new features into simple tutorial videos, making it easier for users to understand and get started.
YouArt Agent can now make tutorial videos. Make every feature you launch a tutorial video. Book a demo to learn more.
44
44
74
22,122
“Tidy the living room” leaves a lot unsaid. I’m curious about how CRIS-0 decides what belongs where from the scene in front of it.
This is CRIS-0, Aether AI's first causal robotic intelligence system, operating in a real home. It reasons about how its actions will change the world, checks what actually happened, and adapts when reality doesn't go as expected.
14
36
39
31,225
Before GTA 6, apparently we needed an endless interactive AI livestream where Gollum fights Duo and the audience can support a fighter to charge an ultimate. I can't argue that this was the missing piece of civilization, but I do want to know who you would back. The stream is Meme Arena, powered by Vivix W1; head in and give your side some support. Choose carefully. You are volunteering to care about the outcome of an owl-versus-Gollum fight.
We broke AI video’s quality–speed–cost trade-off. And turned it into an endless livestream you can play. Introducing Meme Arena, powered by Vivix W1, our real-time interactive video model. High-quality video. Real-time generation. Dramatically lower costs. Together, they make endless interactive livestreams affordable enough to keep running. Now put W1 to the test: 1️⃣ Turn an image into a fighter, or a weapon. Add a few words, and bring your wildest creation to life. 2️⃣ Win the bid. Pick your fighter. Arm them. Watch W1 generate the fight live. 3️⃣ Fuel their ultimate. Support either fighter to build their charge. 4️⃣ Win. Stay. Fight again. The champion faces the next challenger. W1 keeps generating. How long can your fighter hold the arena? Join the livestream: platform.vivix.ai/playground…
55
40
75
22,786
understanding + reasoning + voice getting closer together is where this all goes. looks like they know it
We’re excited to announce that @ultravox_dot_ai is now part of Inworld. Ultravox is the platform developers use to build real-time voice agents. We've worked with the team for a while through our TTS partnership, and today members of the team that built it are joining Inworld to keep developing it.
4
24
28
20,845
Visual coding looks simple on paper: see → understand → build → inspect → fix → repeat. The difficult part is getting AI agents to execute that entire loop consistently and accurately. That’s exactly what PPTBench puts to the test across 18,000 reconstruction tasks.
The next test for AI agents is not just whether they can write code. It is whether they can look, understand, code, inspect—and fix. Today we’re releasing PPTBench, a benchmark for Visual Coding through scientific diagram slides reconstruction. 500 tasks. 36 configurations. 18,000 reconstructions. 64.03% failed the semantic check. Only 2.57% passed all gates with no recorded defect. Visual coding is still wide open. 🤓 Read agent sessions on AgentGit: agent-git.com/@einsia/ppt-be… Full breakdown 👇 Project: lab.einsia.ai/pptbench Paper: arxiv.org/abs/2609.29718 Github: github.com/Einsia/PPTBench
6
16
20
16,892
Stop letting your best agent work get lost deep inside a chat window. 🛑 Start free → showly.ai Your AI agent is already crushing it by writing reports, digging into research, and building sites. The problem? All that effort usually dies as an endless chat response, a clunky .md file, or a localhost page that nobody else can even access. Enter Showly, an Agent Artifact Cloud. Just tell your agent to output its work as a web page, check it out in a private review, and then drop a single link whenever you are ready. The agent handles the creation. Showly handles the cloud delivery, complete with version history and customizable access limits — keep it private, lock it with a password, restrict it to specific emails or your company domain, or open it up to the public. 🔒 It plays perfectly with Claude Code, Cursor, Codex, OpenClaw, and Hermes. Zero need to learn a brand-new editor. ⚡ Go straight from raw agent output to a fully shareable artifact. 🎯
4
26
29
16,223
309B total and 15.5B active is one of those numbers where you have to stop and read the architecture section instead of just reading the headline.
Introducing Naive-N0.5-Flash: Building Frontier AI with AI 🔹 309B MoE, 15.5B active: top-tier in coding, leading in AI R&D. 🔹 Native 1M context, no full-attention layers (hybrid SWA + DSA). 🔹 Inference runtime built by AI: up to 2,000 tok/s in Ultrafast mode. Weights are open today under MIT license. 🔗 Tech blog: naive.ai/en/research/ 🤗 Hugging Face: huggingface.co/NaiveAI/Naive… 💻 GitHub: github.com/NaiveAI-Labs/Naiv… 🌐 naive.ai/en/
3
14
18
16,673
There’s a big difference between generating an ad and producing an ad. Generating gives you footage. Producing means deciding what the shots should be, how they connect, how the pacing feels and what needs to change. GPT Image 2.5 is coming to CapCut and will be available through Design Studio and AI Image. I used it to build the storyboard for an iPhone Duo commercial. Then CapCut PC + Seedance 2.5 took it into video production, with editing and AI-powered refinement after generation. That’s the workflow I find more interesting. #GPTImage25 #CapCutPC #AIVideo #Seedance25
11
156
158
24,190