DeepMind and Harvard just created a virtual rat powered by an AI neural network.

It's capable of mimicking agile movements and neural activity of real-life rats — massively opening up new research with testing on AI animals, and expanding robotics

OpenAI CTO Mira Murati just said that the company's models are ‘not that far ahead’ of what the public has for free right now.

This means OpenAI doesn’t have some GPT-5 level model waiting in the wings... Or is it just de-emphasize before the big reveal?

Synthflow, an AI startup building conversational AI agents a new multilingual update to the platform.

AI call assistants can now communicate in nearly any language, in any voice

Researchers from Microsoft, Google, MIT, and Oxford released new research presenting DenseAV.

It's an AI algorithm that learns the meaning of words and sounds by watching videos, and can help better understand animal communication and discover language.

Elon Musk officially withdrew his lawsuit against OpenAI.

If you missed it, the previous filing accused OpenAI of abandoning its mission of ‘developing AI for the benefit of humanity’.

Still feels like a feud that is far from over...

Zeta Labs came out of stealth, a new AI agent startup built by former Meta Engineers.

The startup unveiled 'Jace,' an AI agent capable of executing complex browser tasks autonomously, similar to CognitionLabs' Devin.

Stability AI released the open model weights for Stable Diffusion 3 Medium.

The 2B parameter text-to-image model offers "advanced photorealism, prompt understanding, and typography capabilities"

Apparate Labs launched PROTEUS, a new real-time AI video generation model.

It creates realistic avatars and lip-syncs from a single reference image, similar to VASA-1, but it's completely real-time.

AI video is accelerating

Runway just announced Gen-3 Alpha: their video generation model to answer Dream Machine and KLING last week

It’s capable of generating highly realistic, 10 second clips from text, or images

Google DeepMind just shared progress on their new video-to-audio (V2A) tech

Until now, AI video generations have been silent, this solves that. V2A can generate an "unlimited number" of tracks for any video.

Here are some thoughts & examples (Sound up 🔉 )

What's cool about this tech is that it's pairable with AI video generators, but we can also use it on any existing video

Meaning we'd also be able to generate audio for any silent films from the past, or any video with broken audio

There's a ton of creative opportunity here

Another thing that stands out is that V2A can understand raw pixels

Unlike other text-to-sound generators, adding a text prompt is optional, because it can understand what's going on in the video

BUT this also means the quality of audio is dependent on the quality of video

OpenAI and Color Health just announced a partnership to create an AI assistant to craft personalized cancer care

The AI (built with GPT-4o) analyzes patient data, guidelines, and medical records to identify screening gaps and create tailored plans.

Tik Tok just launched Symphony, a new suite of AI features including digital avatars, translation tools, an AI assistant, and more.

Brands can choose from a selection of ‘stock avatars’ based on real actors OR create custom avatars to serve as virtual brand reps 🤯

TikTok also announced "Translate for global reach"

It's a new AI Dubbing tool that automatically transcribes, translates, and dubs videos into 10+ languages, helping brands scale content globally.

Similar to Mr.Beast, at scale

Meta's Fundamental AI Research (FAIR) group just published an array of new open-source AI models and techniques.

Some releases include multimodal language tasks, text-to-music and audio, detecting synthetic speech, and more.