Step 7: What to Work On
The general theory for Squid Labs was that we'd take consulting and prototype-building projects, and funnel that money into a series of internal project ideas. If any of the internal projects met with success, we'd license the technology or spin the technology out into a separate company and find people to run that company while we stayed focused on early-stage innovations.
The consulting and prototyping projects initially came through our various contacts from the Media Lab. The Media Lab is unique among graduate programs at MIT, and probably even more generally unique, in that the lab is sponsored by various corporations and governmental agencies and the graduate students are expected to demonstrate and show off their group's work to the sponsors on a very regular basis. On average, I gave one half-hour demo per week to a wide variety of people ranging from Hallmark to DARPA. With that much practice, it's hard not to get really good at telling a compelling story and thinking on your feet. And, the sponsors aren't dumb: good students typically received a job offer of some sort at least every month. It was these opportunities and contacts that we converted into our initial consulting work.
Some how-to advice amidst a self-indulgent story: To successfully start a company or market yourself to attract clients, you'll need to practice telling your story. If you don't already have multiple opportunities to practice, create some by joining a local group that holds public events (like dorkbot or Maker Faire) or putting a sign on your door that says "come in for a tour of my really sweet projects." Local museums, colleges, and universities may have interesting clubs you can join or visit. Put projects up on Instructables. Practice telling your story to both experts and the general public -- these are two very different skills, and you should learn both.
In my experience, people who are good at making stuff (by machining, coding, CAD, designing circuit boards, etc...) are always in demand, because there are fewer of us than people who can't make stuff. Once others know you can make stuff and that you're looking for side-projects or consulting work, opportunities should follow -- even without a Media Lab network to leverage.
The internal Squid Labs projects came from various ideas we'd been thinking about and working on outside of our official graduate research projects. These included battery vending machines, electro-spinning of nanoparticle fibers, ThinkCycle (a precursor to Instructables), tons of carbon nanotube ideas, flexible membrane molding for eye-glasses, Howtoons, kites and kiteboards of all descriptions, modular bikes, and strain-sensing rope. To give credit where credit is due: Saul is a total master of taking the first steps on a project and getting people excited; most of the above list started with him. We use to joke that Saul would start projects, and I would finish them. (Keep this in mind when choosing partners, as complementary traits are a bonus.)
There were a few things we specifically choose not to work on, despite existing opportunities. Chief among them were any technology that MIT had rights to. Colin, Saul, and I had all been a part of licensing technology that we had developed within MIT to a company that Colin and some others from our lab were founding. MIT's technology licensing office and the Media Lab itself both did some things that were counterproductive to the success of the young company and the happiness of the people involved. They made the process so painful that we vowed to avoid licensing technology from MIT in the future.
As an engineer, I feel that my strength is "problem solving" rather than deep knowledge of a specific technique or technology. There are pluses and minuses, but this is definitely the right attitude for attacking problems I've never thought about, and it helps when projects that aren't gaining ground need to be dropped.