Self-Replicating Machines, Open Source Hardware, and the Relationship Between Information and Structure


August 8, 2005


Saul Griffith




We have just survived a brief period in history where we all disembodied ourselves from the physical world and focused on the digital and the informational. It was definitely fun times for many, but more interesting is the way the experience allows us to re-engage with the physical world now that we are on the other side of it. Atoms and bits need not be strangers, and looking at the physical world with a mind to how bits describe it looks like it might be a pretty interesting new perspective. This talk will wander across some of the ramifications of looking at structure as information – physically self-replicating machines, open source hardware, file sharing for hardware hobbyists, and why the hottest tool in your garage will be connected to your PC.


Saul Griffith

Dr. Saul Griffith is an MIT alumnus with multiple degrees in Materials Engineering and Mechanical Engineering. Saul completed his PhD at the MIT Media Laboratory in 2004 on self-replicating hardware and the role and limits of information and state in the self-assembly of complex structure. While at MIT, Saul co-founded Low Cost Eyeglasses, a company using two novel technologies to provide prescription eye-care at low cost for rural and developing communities. Also at MIT, Saul started Howtoons with Joost Bonsen and Nick Dragotta. Howtoons are an alternative curricular for hands-on-science and engineering, illustrated in playful cartoons. A deep interest in the use of social networks for engineering and design led Saul to co-found Thinkcycle and Instructables, experimental platforms for enabling Open-Source approaches to developing physical objects.Saul’s principal research focus is in new multifunctional materials and in minimum and constrained energy surfaces for novel manufacturing techniques. If any description were to tie this seemingly broad array of interests together it is that the last 40 years of developments in logic theory and software and documentation enable new ways to look at the way we build and manufacture things. Why, for example, can’t physical objects have source code? If we think of the physical elements of a machine as parts of a program, how do we utilize physics and information to define the resulting objects that the machine/s produce? Why not teach children the elements of logic and programming using hands-on physical exercises rather than computers? Why not consider the various processes used in making things as shareable sub-routines in a greater library of manufacturing?Saul received the $30,000 Lemelson-MIT award for inventiveness and was awarded the Collegiate Inventors award by the National Inventors Hall of Fame. He has received numerous other awards in design and engineering. Saul holds multiple patents and patents pending in textiles, optics, &