I saw today that in the journal “Lab on A Chip,” Rhee and Burns published a new design for modular microfluidics. Microfluidics has been my life for the last 5 years. I think I’ve mentioned it at some point. It’s been an interesting way to go about science and I’m glad to have been doing it. I can see how lots of projects would be easier if people knew how to use these techniques.
It’s a lot like programming, actually. If you have a problem in the digital world, and you solve it with a clever program, then you’re good to go. It’s easy to repeat it, and you can share the design easily, and the next person who uses it doesn’t have to learn the same level of skill. That’s key: once a programmer gets something to work, it’s a program. The next person just has to run it.
I don’t know if the magic of that is clear to people. Imagine if you were a blacksmith. You train for ten years, build your shoulders, learn the dark luminous secrets of molten iron. Then you can make amazing things like the gate to the winter palace in St. Petersburg. Now let’s say you want to be able to share that ability. You can’t just post it on the ‘net. You can share some ideas, maybe a 10 year curriculum that would help develop the skills… but the skills are not transferrable.
These days, if you have an idea and you write it into code, and you post it on the net, anyone can do what you did. With a click. No practice is required. But what about other, more physical things? In the next while, if Gershenfeld is to be believed, we are going to see material things produced by open source software. The RepRap project is gaining some momentum already. But in the microfluidics arena, a certain kind of open source physical goods is already there.
People publish designs and those designs can be reproduced by people who have only limited training in things like fluid mechanics, lithography, and cell culture. Once produced, they open whole avenues toward the data that was once only obtainable by people with years of skill and training. And it will only get better.
How does modular microfluidics fit into this? That’s another step toward anyone being able to build these devices. A number of user facilities will generate the master for replication molding. Once generated, that master can be used to produce hundreds of the modules that the paper describes. Once produced, these modules are like toy bricks: they can be used to produce anything, from automated, computer controlled chromatographs to microeractors.
I suspect that in 10-20 years, the complex synthesis for all kinds of substances will be reduced to a set of a few of these blocks (or something like them). I can imagine that, in principle, anybody could take a simple instruction set, have their RepRap print it, hook it up to their computer and have it produce LSD from a few household chemicals.
How will that play out, legally and socially?