Ai Agents

I used a movement sheet as a reference image to animate the dance using Seedance 2.0 + GPT image 2.0 GPT Image 2.0 prompt: [STYLE] Monochrome grayscale illustration, 3D-rendered character, clean instructional reference sheet, white background, comic-style cell grid layout, technical diagram aesthetic. [LAYOUT] 4×4 grid layout with a total of 16 panels. Each panel is separated by thin black border lines. Cells are numbered from 1 to 16, with consistent panel sizes. [CHARACTER] image1 (the same character appears consistently in all panels) [PANEL STRUCTURE – per cell] Top-left: bold number badge + English title text Center: full-body character pose illustration Bottom-left: English description text (3–4 lines) Overlay: directional arrows indicating movement [ARROWS / MOTION INDICATORS] Curved arrows, straight arrows, and circular rotation indicators placed around the character to show motion flow and direction. [RENDERING STYLE] Highly detailed 3D sculpted style, soft studio lighting, subtle shadows, no color, grayscale shading, clean linework, game concept art quality. [NEGATIVE] No background scenery, no color tones, no additional characters, no complex background.

STOP BUILDING ANIMATIONS FROM SCRATCH react bits has 110+ animated react components text effects, backgrounds, 3d interactions reactbits.dev copy-paste or cli install... works with gsap, framer motion, tailwind, js or ts your choice github.com/DavidHDev/reac…

YOU DON’T NEED A LAB TO BUILD HARDWARE ANYMORE Tinkercad lets you design circuits, simulate Arduino code & 3D model all in your browser no installs. no setup. no excuses perfect for beginners → powerful enough for real prototyping

Someone just built a desktop app that that generates 3D models from images and runs 100% locally. It's called Modly. It runs entirely on your GPU, no cloud, no API bills. Just drop an image and get a 3D mesh. 100% Open Source.

someone built a free, open source app that lets you run codex, claude code, code editor, terminal, and your entire dev environment directly from your phone app link: apps.apple.com/us/app/lunel/i…

Cloud GPU training is a scam. A single M4 MacBook does 2.9 TFLOPS. Seven friends with MacBooks match an NVIDIA A100. @AlexanderCodes_ just open-sourced a tool that makes this work over Wi-Fi. It's called AirTrain. Here's how it works: Traditional distributed training (DDP) syncs gradients after every single step. For a 124M parameter model, that's ~500MB exchanged per step. You need 50 GB/s of sustained bandwidth. Impossible over Wi-Fi. AirTrain uses the DiLoCo algorithm. Each Mac trains independently for 500 steps, then syncs only the difference. One sync per 500 steps instead of one per step. 500x less network communication. Wi-Fi actually works. The entire sync takes ~2 seconds. Here's what makes it wild: → Zero-config discovery. Devices find each other automatically via mDNS/Bonjour. Same protocol as AirDrop. → Fault tolerant. Nodes can join and leave mid-training without killing the run. → Checkpoint relay. Train for a few hours, export a checkpoint, hand it off to someone else to continue. Like a relay race for ML training. → Built on Apple's MLX framework. Native to M1/M2/M3/M4/M5 unified memory. No host-to-device copy overhead. → Local dashboard. Real-time loss curves, peer monitoring, throughput metrics in your browser. Here's the wildest part: An M4 Max with 128GB unified memory can train a 70B parameter model without offloading. An NVIDIA RTX 4090 has 24GB VRAM. Apple Silicon gets ~245-460 GFLOPS per watt. Training on MacBooks costs almost nothing in electricity compared to cloud GPUs. And there are hundreds of millions of Apple Silicon Macs in the world. The math: Traditional DDP: 1 sync per step = 50 GB/s required AirTrain (DiLoCo): 1 sync per 500 steps = 0.1 GB/s required Wi-Fi handles 0.1 GB/s. That's it. That's the breakthrough. They even built a community platform at airtrain.dev with live session browsing, checkpoint sharing, and a contributor leaderboard. Training a 124M parameter GPT-2? Instead of renting cloud GPUs at $3/hr, pool three MacBooks in a coffee shop and train for free. MIT licensed. Built in Python. 1 contributor. Early stage but the idea is insane. 100% Open Source. (Link in the comments)

I was paying $229 for Screen Studio. Then I found this repo and uninstalled it the same night. Recordly is a free, MIT licensed screen recorder and editor that ships every feature I was paying for auto-zoom animations, cursor smoothing, motion blur, native macOS capture, and a full timeline editor. → Apple-style zoom animations with automatic suggestions from cursor activity → Native ScreenCaptureKit recording on macOS 12.3+ → Cursor smoothing, motion blur, click bounce, macOS-style cursor assets → Smooth pan transitions between zoom regions → Timeline trim, speed ramps, annotations, manual zoom spans → Wallpapers, gradients, padding, rounded corners, blur, drop shadows → MP4 and GIF export at any aspect ratio → .recordly project files so you can reopen edits later macOS, Windows, Linux. Scene composition runs on PixiJS. Native Swift helpers handle macOS cursor telemetry. 8.3k stars. MIT License. 100% Opensource github.com/webadderall/Re…

A free, open-source tool can now swap your face in real-time on any webcam. One photo. 3 clicks. Works on live video calls. 79,500+ developers starred it on GitHub. Ars Technica called it "truly terrifying." Here's why this changes everything about video call trust: It's called Deep-Live-Cam. You upload a single face image, select your webcam, and hit "Live." Within seconds, your face is replaced in real-time matching your expressions, head movements, and even lighting conditions. It supports mouth masking (so lip sync stays accurate), multi-face mapping (swap different faces on multiple people simultaneously), and GPU acceleration on NVIDIA, AMD, and Apple Silicon. The wildest part? It went so viral that IShowSpeed, SomeOrdinaryGamers, and TechLinked all covered it live. Streamers were swapping into celebrities mid-broadcast and their audiences couldn't tell the difference in real-time. PetaPixel, Creative Bloq, Dataconomy, and DIYPhotography all ran stories on it. The coverage wasn't "cool new AI toy" it was "this is a security problem." Built-in safeguards exist the tool blocks nudity, graphic content, and sensitive material. The developers require consent for using real faces and labeling outputs as deepfakes. But the technology is out there now. Real-time face swapping used to require Hollywood budgets and render farms. Now it runs on a laptop with a consumer GPU. If your company still uses video calls as identity verification, that's a problem that needs solving yesterday. 100% open-source. AGPL-3.0 license.

THIS IS MIND BLOWING. Zhejiang University just open-sourced the first complete GUI agent pipeline that trains, evaluates, AND deploys to real phones. It's called ClawGUI. It got 3 modules and 1 framework. - ClawGUI-RL trains agents on real Android/iOS devices, not just sandboxes - ClawGUI-Eval reproduces results across 6 benchmarks and 11+ models at 95.8% accuracy - ClawGUI-Agent puts trained agents on your phone through Telegram, Slack, Discord, and 9 more platforms Their 2B model outperformed Qwen3-VL-32B. 16x smaller. Better results. The gap between AI research and your actual phone just closed.

I recorded a 25-min tutorial on prompting $50,000 landing pages with Gemini + Nano Banana

8 RAG architectures for AI Engineers: (explained with usage) 1) Naive RAG - Retrieves documents purely based on vector similarity between the query embedding and stored embeddings. - Works best for simple, fact-based queries where direct semantic matching suffices. 2) Multimodal RAG - Handles multiple data types (text, images, audio, etc.) by embedding and retrieving across modalities. - Ideal for cross-modal retrieval tasks like answering a text query with both text and image context. 3) HyDE (Hypothetical Document Embeddings) - Queries are not semantically similar to documents. - This technique generates a hypothetical answer document from the query before retrieval. - Uses this generated document’s embedding to find more relevant real documents. 4) Corrective RAG - Validates retrieved results by comparing them against trusted sources (e.g., web search). - Ensures up-to-date and accurate information, filtering or correcting retrieved content before passing to the LLM. 5) Graph RAG - Converts retrieved content into a knowledge graph to capture relationships and entities. - Enhances reasoning by providing structured context alongside raw text to the LLM. 6) Hybrid RAG - Combines dense vector retrieval with graph-based retrieval in a single pipeline. - Useful when the task requires both unstructured text and structured relational data for richer answers. 7) Adaptive RAG - Dynamically decides if a query requires a simple direct retrieval or a multi-step reasoning chain. - Breaks complex queries into smaller sub-queries for better coverage and accuracy. 8) Agentic RAG - Uses AI agents with planning, reasoning (ReAct, CoT), and memory to orchestrate retrieval from multiple sources. - Best suited for complex workflows that require tool use, external APIs, or combining multiple RAG techniques. 👉 Over to you: Which RAG architecture do you use the most?

Anthropic has quietly dropped a massive curriculum of free courses covering the entire AI ecosystem The syllabus is STACKED 🔥 → Claude Code: CLI automation for your workflow → MCP Mastery: building custom tools and resources in Python → API: a complete guide to the Anthropic backend → AI Fluency: frameworks for safe and efficient collaboration → Claude 101: core features for everyday work I added the link to the free @AnthropicAI Academy in the 🧵↓

@aryanXmahajan Ads

R.I.P Media planners. This AI agent generated $2B+ with Ads — and just killed manual creative research forever... → No more doom-scrolling Meta Ad Library for hours → No more watching competitor videos one by one → No more "inspo folders" you never open again → No more creative briefs built from gut feeling Just one prompt → full competitive intelligence in 60 seconds. Here's how it works: → Category Crawler (scrapes competitor ads, Reddit, search trends automatically) → Visual Ad Dissector (breaks down hooks, pacing, copy, angles, offer framing) → Ad Account Connector (compares what the market is doing vs. what YOU'RE doing) → Gap Finder (identifies winning patterns your brand hasn't tested yet) → Brief Generator (outputs production-ready scripts and creative briefs instantly) Results from teams using it: • 20+ hours of manual research → 60 seconds • Competitors’ ads, found from over 150M+ ads. • Production-ready scripts and briefs — no guesswork It also has a free Competitor Intel report: Enter your website and get competitors' ads, best hooks, winning angles, and gap analysis against your own ads. Like + comment "ADS" + repost, and I'll DM you the link. (must be following)

BREAKING: Someone just built an AI that codes and browses the web at the same time. It's called Accomplish and it runs locally without burning through API credits. No Claude Desktop. No Cursor. No monthly subscriptions. 100% Opensource.

completely Free Habit Tracker :tarli.exploo.xyz

You can now build Claude Code workflows visually 🤯 Managing AI agents in the terminal was a pain. cc-wf-studio lets you visually map out your coding workflows. drag, drop, automate. 100% open source.

🚨BREAKING: Someone built an AI coworker that actually remembers everything you've discussed. It's called Rowboat and it builds a knowledge graph from your work. - Connects Gmail, Calendar, Drive, meeting notes - Runs 100% locally (your data never leaves your machine) - Generates PDFs, briefs, emails from your context - Plain Markdown files you can edit anytime 4.6K stars. 100% Opensource.

@heynavtoor Ai

Everyone is hyped about Claude… but barely anyone knows how to actually use it to replace real work. I collected 700+ mega prompts that turn Claude into a full-blown productivity engine. Comment "AI" and I’ll DM you everything.