HARD MODE
The MIT Hardware × AI Hackathon
Helped organize HARD MODE, a 48-hour hardware × AI hackathon at the MIT Media Lab where 200+ builders prototyped intelligent objects that sense, learn, and respond.
HARD MODE is a 48-hour hardware × AI hackathon at the MIT Media Lab, built around a single question: what happens when AI stops living behind a screen and starts living in physical objects? Over one weekend — March 6–8, 2026 — more than 200 builders in 40-plus teams designed and prototyped intelligent objects: devices that sense, learn, adapt, and respond to the people around them.
The event was hosted by the Media Lab’s AHA (Advancing Humans with AI) program and the MIT Department of Architecture’s Design Intelligence Lab. I was part of the organizing team that put it together.
My role
I helped organize HARD MODE alongside the AHA and Design Intelligence Lab teams — shaping the program, preparing the hardware and fabrication resources participants would build with, and running the floor across the 48 hours. It was one of the most exciting hackathons I’ve been part of: a room full of engineers, designers, and researchers turning models into things you can actually hold.
The format
Rather than a single open brief, HARD MODE was structured around six thematic tracks — the everyday contexts where intelligent hardware might matter most:
- Play — toys, instruments, and playful machines
- Learn — tools that teach the body and the mind
- Work — devices that augment how we make and do
- Connect — objects that mediate human relationships
- Reflect — hardware for awareness and introspection
- Thrive — health, wellbeing, and the body
Teams had access to a full fabrication stack — 3D printers, an electronics lab, sensors and dev kits, and AI compute — and the event was free, with meals, hardware kits, and everything needed to ship a working prototype.
Prizes
- A $50K founder-friendly investment (a SAFE from E14) for the standout project
- $500 per winning team across the six track prizes
- Additional sponsor prizes
What got built
The winning projects pushed AI off the screen and into the body. In the Learn track, a team built Human Operator — a system that guides physical movement through electrical muscle stimulation. It pairs a vision-language model and voice commands (processed through Anthropic’s Claude API) with EMS electrodes to move a user’s hands, fingers, and wrists, helping them perform motions they don’t yet know how to do, from waving to playing piano. The top investment prize went to Dreaming Objects, one of the weekend’s standout teams.
Sponsors
HARD MODE was supported by a broad set of hardware and AI partners, including Anthropic, Akamai, and Qualcomm (platinum); Bambu Lab, E14, IFM, and GigLab (gold); and Hudson River Trading, Seeed Studio, Tenstorrent, Snap, VIAM, BootLoop, Murata, igus, and Nvidia (silver).
Read more at the official HARD MODE site and the MIT Media Lab event page.