MIT Computer Science & Artificial Intelligence Laboratory

A new AI model from MIT CSAIL called "CausVid" crafts smooth, high-quality videos in seconds: bit.ly/3GKkSJQ

The system combines the photorealism of diffusion models w/the speed of autoregressive approaches, potentially making it useful for editing videos & creating training data for robots.

"A good programmer is someone who always looks both ways before crossing a one-way street." — Doug Linder

A new technique from MIT can make AI-generated code adhere to whatever programming language or other format is being used, while remaining error-free: bit.ly/43U2Pua

3D modeling you can feel MIT CSAIL’s “TactStyle” system uses image prompts to replicate both the visual appearance and touch-based properties of 3D models. It could help create home décor, personal accessories, and tactile le...

3D modeling tools dictate how many objects look — but what about how they feel?

MIT's "TactStyle" program enables creators to use image prompts to stylize the textures of 3D models. It can customize the tactile properties of accessories, decor, & more: bit.ly/42HJwBQ

For this computer scientist, MIT Open Learning was the start of a life-changing journey MIT Research Scientist Ana Trišović went from a student downloading MIT Open Learning resources in Serbia to becoming a computer scientist at CERN, Harvard, and MIT.

Ana Trišović's educational journey has been nothing short of transformative! She began learning Python using MIT OCW in 2012 from her home in Serbia, eventually leading to her current position at the FutureTech Lab at @csail.mit.edu

news.mit.edu/2025/mit-ope...

Cable-driven mechanisms enable streamlined, precise motion, but are difficult to assemble by hand.

MIT’s "Xstrings" 3D printing approach automates how these objects are put together, allowing users to customize designs for things like:
- A lizard robot
- A wall sculpture

Full video: bit.ly/4il1aCj

Like human brains, large language models reason about diverse data in a general way MIT researchers find large language models process diverse types of data, like different languages, audio inputs, images, etc., similarly to how humans reason about complex problems. Like humans, LLMs...

Like human brains, large language models reason about diverse data in a general way.

A new study shows LLMs represent different data types based on their underlying meaning & reason about data in their dominant language: bit.ly/3QrZvyy

MIT prof. & CSAIL principal investigator Kaiming He discusses AI’s role in lowering barriers between scientific fields & fostering collaboration across scientific disciplines.

He says that these walls are breaking down because of a "common language that has been found in AI": bit.ly/4hwPXxX

Exciting fabrication projects coming up from CSAIL's HCI Engineering Group 👀 Stay tuned!

A new printmaking technique produces images that can change colors when heated up ♨️

MIT researchers helped develop "Thermochromorph" to take advantage of full-color process printing. It uses thermochromic inks to make images transition into each other: bit.ly/42Jeyeq

Thoughts on DeepSeek-R1 so far?

What are your goals as a computer scientist this year?

To help robots execute open-ended, multi-step chores, MIT CSAIL researchers used vision models to see what’s near the machine & model its constraints. An LLM sketches up a plan that’s checked in a simulator to ensure it’s safe & realistic, potentially aiding household robots: bit.ly/3ZNJWXj

You’ve likely imitated the sound of a cow or a bell before, AKA "sonically sketching."

Inspired by the cognitive science of how we communicate, an AI system from MIT CSAIL can produce human-like vocal imitations w/no training & w/o ever having "heard" a human vocal impression before: bit.ly/3VoJLz8

Another fall day flying by at CSAIL.

v/Mike Grimmett

Hello, world!

Fun fact: This term was coined in Brian Kernighan and Dennis Ritchie's “C Programming Language” textbook, published in 1978.

go.aws/37TmOZq