Vannarot Roeung
Recent posts by Vannarot Roeung
Hume TADA: Open-Source TTS With Zero Hallucinations
Hume AI has released TADA (Text-Acoustic Dual-Alignment), an open-source text-to-speech model that eliminates a persistent problem in AI speech synthesis: hallucinated words
Adobe Launches AI Assistant for Photoshop and Expands Firefly
Adobe launched a public beta of its AI assistant for Photoshop on March 10, 2026, while simultaneously expanding Firefly with generative editing tools and integrations with 25+ third-party AI models including Runway, OpenAI, and Black Forest Labs. Adobe shipped two major updates on the same day
Meta Acquires Moltbook, the Social Network for AI Agents
Meta has acquired Moltbook, a Reddit-style social network built exclusively for AI agents. The platform launched in late January 2026 as an experimental "third space" where AI bots can autonomously post, comment, and upvote or downvote content
NVIDIA NemoClaw: Open-Source AI Agents at GTC 2026
NVIDIA is launching NemoClaw, an open-source enterprise AI agent platform, ahead of GTC 2026 (March 16-19). The platform integrates with the NeMo framework and Nemotron models, backed by partnerships with Salesforce, Google, Adobe, Cisco, and CrowdStrike
Freepik 5.6 Adds Cinematic Shot and AI Music Generator
Freepik has shipped version 5.6 of its mobile app, adding two significant creative tools: Cinematic Shot for generating studio-grade photorealistic images and an AI Music Generator that produces royalty-free tracks on demand
Black Forest Labs Releases Self-Flow: 2.8x Faster AI Model Training
Black Forest Labs, the team behind the FLUX image generation models, published Self-Flow on March 10, 2026: an open-source framework that trains image, video, and audio generation models 2.8x faster without requiring external encoders
Fish Audio S2: Open-Source TTS Beats GPT-4o
Fish Audio released S2, an open-source text-to-speech model trained on over 10 million hours of audio data spanning 50 languages. The model beats GPT-4o-mini-tts on the EmergentTTS-Eval benchmark with an 81.88% win rate, and ships with full weights, fine-tuning code, and a streaming inference engine
Picsart AI Playground Puts 90+ Models in One Creative Hub
Picsart just launched AI Playground, a unified creative hub that brings together more than 90 AI models from 24 providers into a single interface
Runway Characters Turns Any Image Into a Live AI Avatar
Runway launched Runway Characters on March 9, 2026, a real-time video agent API that turns any image into a fully expressive, conversational digital persona
Framer Server API Lets AI Agents Build and Publish Sites
Framer launches Server API in open beta, enabling AI agents and automated workflows to sync CMS, publish changes, and update designs programmatically.
Indie Artists Sue Google Over Lyria 3 AI Music Training
Independent musicians filed a class action lawsuit against Google, alleging Lyria 3 was trained on 44 million YouTube clips without permission.
Claude Code Review Sends AI Agent Teams on Every PR
Anthropic launched Claude Code Review on March 9, 2026, a multi-agent system that dispatches teams of AI reviewers on every pull request. The tool uses color-coded severity ratings and boosted thorough code reviews from 16% to 54% in internal testing
Tripo P1.0 Ships Production-Grade 3D AI at GDC
Tripo debuted P1.0, a native 3D diffusion model that generates production-ready 3D assets in roughly two seconds, at GDC 2026 in San Francisco
LeCun's AMI Labs Raises $1B to Build World Models
Yann LeCun's AMI Labs raised $1.03 billion at a $3.5 billion pre-money valuation, making it Europe's largest seed round ever. LeCun left Meta at the end of 2025 to build "world models" that learn from physical reality rather than text alone
OpenAI's Pentagon Deal: What AI Ethics Mean for Creators
OpenAI signed a military contract that its rival Anthropic refused to accept. The deal triggered the most senior resignation in OpenAI's recent history, an unprecedented government blacklisting of an American AI company, and a legal battle that is still unfolding
OpenAI Robotics Chief Resigns Over Pentagon Deal
Caitlin Kalinowski, OpenAI's head of robotics and consumer hardware, resigned on March 7, 2026, over concerns about the company's recently announced Pentagon deal
OpenAI's $110B Amazon Deal and What It Means for Creators
OpenAI just raised more money in a single round than most countries spend on defense. The $110 billion deal, backed by Amazon, Nvidia, and SoftBank, values the company at $730 billion and reshapes the competitive landscape for every creator who depends on AI tools
Pika AI Selves Lets Creators Build Persistent Video Avatars
Pika Labs launched AI Selves on March 7, 2026, a feature that creates persistent AI video avatars from user-uploaded reference material. Each AI Self has memory, personality traits, and can appear in any scene the creator generates through Pika's existing text-to-video pipeline
ElevenLabs Opens Iconic Voice Marketplace With Michael Caine
ElevenLabs launched its Iconic Voice Marketplace in early March 2026, a licensing platform where celebrity voices are made available to businesses and creators for commercial use
Anthropic Launches Claude Marketplace for Enterprise
Anthropic launched the Claude Marketplace on March 6, 2026, a procurement platform where enterprises can spend their existing Anthropic commitments on third-party applications built on Claude
IBM Granite 4.0 Speech Tops OpenASR Leaderboard
IBM released Granite 4.0 1B Speech on March 6, a compact multilingual speech model that now ranks first on the OpenASR Leaderboard for automatic speech recognition
Luma Agents: What Multi-Model AI Means for Creative Studios
A $15 million, year-long advertising campaign reduced to localized multi-market ads in 40 hours for under $20,000. That is the headline number from Luma's new Creative AI Agents, launched March 5 with enterprise clients already running production workloads
LTX-2.3: Open-Source AI Video With Desktop Editor
Lightricks released LTX-2.3 on March 5, 2026, a 22-billion-parameter open-source video generation model with native audio, portrait video support, and a companion desktop editor that runs the entire model locally. The release ships under Apache 2.0, making it free for commercial use and fine-tuning
Modular Diffusers: Composable AI Image Pipelines
Hugging Face launched Modular Diffusers on March 5, replacing the monolithic DiffusionPipeline with a composable block system