Once marshland and farmland in the 1700s, the corner of Cambridge where MIT’s Stata Center now stands became a center of innovation during World War II with the construction of MIT’s legendary Building 20, a temporary structure that housed radar research and later launched some of the university’s most iconic ideas in everything from linguistics to robotics.
Today, that same spirit of experimentation lives on in the Stata Center, a bold, modern building designed by famous architect Frank Gehry. Built in 2004, it now houses an outpost of MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL), where some of the most ambitious minds are rethinking how humans and machines can co-create physical objects through shared design language—not just commands.
MIT Stata Center. Image courtesy of 3DPrint.com.
During my visit to the Human-Computer Interaction (HCI) Engineering Group, led by Professor Stefanie Mueller, I met with two of her researchers: Faraz Faruqi and Jiaji Li. Both are deeply immersed in rethinking how everyday people (not just engineers) can design, modify, and fabricate physical objects with the help of AI.
A Different Kind of Tool
At first, the objects on the lab tables—pillboxes, phone stands, utensil holders—look like everyday items. But if you take a closer look, there’s something different. Each one was designed or customized using AI tools developed right in the lab. These tools let users change the look or style of a 3D model while still making sure it works, like keeping a pillbox easy to open and close.
“Right now, most assistive devices come in bland, standard designs, and they don’t reflect the people who use them,” Li explained as he showed me a flower-patterned utensil grip, one of many 3D printed applications. “Our goal is to change that. We want to give people the power to personalize these tools, not just aesthetically, but in ways that preserve and even enhance their function. That’s the real power of design.”
This blending of function and flair is where CSAIL’s HCI group shines. One tool, Style2Fab, allows users to apply aesthetic changes to 3D models without disrupting the parts that matter most, like screw threads, hinges, or structural supports. For example, a phone stand can be stylized with organic patterns or colorful finishes, while still holding your phone securely.
The AI behind these tools doesn’t just make pretty shapes; it understands what not to change, and that’s what sets it apart.
Faruqi and Li used Style2Fab to create dozens of items. Image courtesy of 3DPrint.com.
A Collaborative Future
“AI shouldn’t replace the designer, it should support them,” Faruqi told me. “This is a core theme of our work. The goal is to reduce the time it takes to create something functional while still keeping people in control of the process.”
He showed me a complex model, something that used to take a week to manually design in Rhino and Grasshopper. “This used to take me days. Now, with the tools we’ve built, it takes minutes,” he said. The system even lets users set preferences, such as how flexible or rigid an area should be, and encodes those properties directly into the model.
“A lot of assistive products are functional but dull,” Li said. “We’re trying to make them feel more personal, something people are proud to use.”
Faruqi and Li are part of a growing movement to democratize digital fabrication. They envision a world where anyone, not just engineers or industrial designers, can co-create with machines. That vision includes making it possible to generate new objects from scratch using text prompts, while the system predicts and tests how that object will perform in the real world, before it’s ever printed.
“I’m more excited now about generating 3D models that actually encode physical properties,” he said. “That way, you can predict how something will perform before you even print it—and loop that feedback into the design process. The idea isn’t that the machine does everything, or that the human does. It’s about designing the interaction together.”
I’d spoken with Faruqi more than once last year to discuss his research at CSAIL, but seeing it come to life in person, surrounded by the very tools and prototypes he described, was something else entirely. Although the lab was quiet for summer break, there was still plenty happening. Traces of ambitious projects filled just about every corner of the HCI space, from adaptive tools designed for people with mobility challenges to large-format 3D prints that blended design, engineering, and material experimentation.
Inside HCI, software and hardware are not treated as separate domains, but as co-creators in what Li described as “a fluid design process.” And this philosophy isn’t isolated to one lab. Across CSAIL, there’s growing momentum around integrating intelligent systems with designing and producing real things. For example, one wing is being transformed into a dedicated robotic AI fabrication lab, a sign that this vision is expanding.
MIT Stata Center. Image courtesy of 3DPrint.com.
Objects with a Story
At the HCI lab, I saw some cable-driven 3D printed prototypes: some resembled tentacles, swinging and grasping as in-place-printed cables are pulled; others looked like lizards, crawling across the ground through cable actuation by a motor. They merged biomimicry with robotics, while their motion was studied using the Xstrings design tool to explore how material threading, tendon placement, and interaction design could come together in a seamless, assembly-free 3D printing.
“I want to print various mechanisms directly into objects—not just the shapes, but also the ways they move and interact,” Li told me. “Ultimately, I hope that 3D-printed objects can crawl out of the printer like living creatures.”
From left: Jiaji Li, Vanesa Listek, and Faraz Faruqi with the 3D printed items. Image courtesy of 3DPrint.com.
Faruqi also showed me a 3D printed self-watering planter, made with their Style2Fab tool. The inner water system was left unchanged, so the planter still worked, while the outside was free to reflect personal style.
These were all examples of a bigger goal: helping people design objects that aren’t just visually interesting, but actually interactive and useful. The HCI lab is full of stories; I only scratched the surface during my visit.
As Faruqi pointed out: “Fabrication doesn’t have to be a tedious process. It can be as simple as talking to your machine about the interaction you want to have—then figuring out the solution together. It’s not just the machine or the human. They work as a team.”