When Cameron Sinclair took the stage to receive his TED Prize last Thursday, he devoted a good portion of his talk to the exploration of his idea for an "open source" form of architecture. Cameron's emphasis was on the openness of the designs and architectural innovations most useful to builders in the developing world, but the idea of open source architecture has the potential to go even further than that. It dovetails with the slow emergence of open source hardware, pointing us towards a world of individual power over design that has the potential to be extraordinarily worldchanging.
At first blush, the open source model doesn't necessarily seem like a good fit for physical objects. The reason comes down to replication: with open source software, I can give you a copy of (for example) Firefox without losing my own. With open source hardware, at present the best I can give you a copy of the instructions as to how to make your own (for example) microprocessor with your own equipment; it's the functional equivalent of saying "you want Firefox? Here's the source code -- go find a compiler and learn how to read and modify the code to work with your hardware." While this is hardly a barrier for people who already do have the necessary tools and skills, it's a major hurdle if open source hardware is ever to move beyond a few niche fields.
But replicability is something of a secondary effect of open source -- if you give someone a copy of a program's source code, it's hard to prevent them from giving away complete versions. The key elements of open source, accessibility and modifiability of the underlying instructions, apply just as readily to hardware designs as they do to software code. Moreover, the lack of replicability is something of an artifact of technology; that I can't send you a copy of a computer chip (or mobile phone, or fabber, or bacteria), only a design, is important only in that you can't just double click something and see that microchip (etc.) pop up in front of you.
As fabrication technologies take on an increasingly digital form, this distinction will become less clear. When I send you a copy of Firefox, what I'm actually sending is a set of instructions telling your computer how to organize bits of magnetic material on a hard drive to create another instantiation of Firefox. Similarly, at some point in the next decade or two, I will be able to send you a "copy" of my phone simply by sending a set of instructions telling your computer how to organize bits of carbon in a desktop nanofactory to create another instantiation of a phone. Very soon, a more software-like open source physical world paradigm will become possible.
But what will really make a big difference will be the emergence of tools allowing you to take that set of instructions for the phone and modify it to meet your particular needs prior to it being printed out.
I don't mean a requirement that you design your own phone from basic parts on up, or can only get a bare-bones phone that you then must seek out and add on (potentially conflicting) modules to give it any kind of real utility. I'm talking instead about an interface to the code that makes it easy for amateur designers to tweak the instructions without running the risk of easily breaking the final result. In effect, I imagine that we'll soon be in a world of "skins," "plug-ins," and "overlays" for the material world.
Download any application aimed at an audience of teenagers or young adults, and you'll be confronted by a plethora of options for making it work and look the way you want. "Skins" for programs change the external appearance in sometimes dramatic ways, with (usually) little significant modification of the interface; "plug-ins" are (generally) the reverse, altering the operation of the software, but doing little to change its look and feel; "overlays" are add-ons intended to provide a specific change to the visual use of a program, but (typically) not to change how the whole application works or looks. In nearly every case, programmers code the programs explicitly to allow (or even encourage) these kinds of add-ons. And in most cases, the add-ons are open source.
It's possible to make Firefox look and behave in ways that would make it seem utterly alien to a user of the unmodified version. What's more, many of the modules that enable these changes can be built by people with little or no programming experience. The open source code for Firefox enabled the creation of tools allowing the easy creation of add-ons, while shielding the core application code from seriously disruptive modifications. The open source code of most of the modules enables new users to learn-by-copying, and to modify near-miss designs to meet specific needs.
So imagine a world where these hand-holding tools allow changes to the design of physical goods, not just software. You want that CO2-analysis module on your phone? Drag-and-drop a plug-in to the phone code and press "print," and the underlying code handles how to fit the pieces together. Or, from an architectural perspective, you want an extra space in that clinic to serve as a classroom? Drag the room plug-in around the 3D design until you get the right location, then let the code deal with issues of load-bearing walls and routing power outlets. If the plug-in modules are simple enough, you can even go in and tweak the parameters to get the results you need.
I suspect that professional product or architectural designers are already sharpening their knives -- clearly, this is a scenario in which everyday citizens don't necessarily need the full range of services from designers to accomplish some otherwise moderately sophisticated tasks. But just as it's possible but not desirable to use the tools to modify Firefox in order to use it as a word processor, these "plug-ins" for physical objects wouldn't be a substitute for dedicated design. Furthermore, someone has to come up with the code for how the plug-ins work in the first place -- this scenario doesn't eliminate designers, just moves them up the food chain a bit.
And while the nanofactory aspect of this scenario is off in the future, the core idea is something that could work right now. When Cameron's open source architecture database is built, it should be possible to give it an interface which would allow non-architects to modify the designs safely to meet particular demands.
As we've observed numerous times before, one of the key advantages to an open source approach is that it allows people to examine the code as a tool for education. As the proliferation of skins, plug-ins and overlays for a huge variety of programs demonstrates, the addition of an easy-to-use interface for coders can open up software design to a larger variety of people than those willing to engage with C++. There's no reason why these two principles can't be part of an open source design and open source hardware world, too.
quick example to point out. there seem to be pretty clear similarities here to the development of automation technology in the US, in terms of moving fundamental procedures out of the hands of pre-existing industrial processes.
now, david noble has a wonderful book about this development in machine tools, and he argues with much evidence that the particular character of automation in the US was organized in such a way as to reduce the actual shopfloor power of skilled workers and break their unions. importantly, he doesn't argue that automation itself would necessarily do this. other automation technologies in development at the time yielded good results but left a great deal of authority in the hands of labor. in fact, one prominent engineer, a New Deal Democrat, designed a model explicitly to keep a great deal of shopfloor power in the hands of employees. he approached the unions and asked them to support this technology to stave off the management technologies developed with the military. the unions weren't interested, and soon after the machine tool sector was automated and their unions functionally smashed.
yet what we find in the US today is a crippled machine tool sector, even though it's automated, and we find generally weak manufacturing. compared to, for instance, that of germany or sweden. rather than simply crush the skilled workforce, in sweden they recognized that the best way to implement automation technology was to pair skilled laborers with computer workers, combining the skills rather than simply trying to replace one set of knowledge with another carte blanc.
i mention this example because of your comments about professional designers, which i think might also apply to a wider range of workers in the world. there is a high road for this and a low road; the low road looks good at first but never lasts, and the high road always is the more sustainable.
i deeply believe in open source values, but in this age we too easily envision technical utopias cut off from human skills and scales. i don't believe we're really talking about open source in "material" things if we displace the lived knowledge of practitioners working with materials and offer instead customers such simple options as "skins". there is an elegant beauty to open source, and it seems like part of that beauty is that it encourages an intimate, artisanl sensitivity and curiosity to a work, to materials and processes. as a knowledge form, it encourages a sort of gentle, quiet understanding of people truly engaged with another thing and working with the materials. if we abandon that level of give and take that is the province of the craftsperson, skilled worker, or practiced designer, that open concern with material, we lose something truly beautiful, something that allows human artifice to become elevated to artistry.
this isn't meant as an argument against this line of development- as i mentioned in the example, the best work integrates that level of skill and respect with new opportunities for design and broader participation. i just mean to re-emphasize what i think was your point, that such developments are made stronger by maintaining an open dialogue with skilled practitioners. this allows us to maintain a dignity and a beauty throughout invention of the new.
Are "David F. Noble" and "David W. Noble" the same person? Many of the bookselling sites seem to think they are.
On the assumption that they are the same person (And please note I haven't read any of his work.) I don't think the comparison Donald made between open source designs of fabricated hardware and the emergence of automated machine tools after World War II is an accurate one.
My own experience, as a self-taught technician working with open source, is that it is profoundly empowering. At least it is for people who are willing to study programming languages and take the time to write the code and install or build some extensions or applications. That's really the only barrier.
I don't know for certain, but my guess is that will continue to hold true for open source hardware designs in a world of ubiquitous desktop fabricators and garage minifactories.
My guess is that what's going to happen is a new kind of hobby will spring up. In additional to all the self-taught, hobbyist woodworkers, carpenters, plumbers, welders, tailors and machinists, there'll be hobbyist fabbers sharing their GPL'd designs and extensions around the world.
How is that an abandonment of the "give and take that is the province of the craftsperson?" It appears to me that, if anything, it democritizes it.
One of the most interesting reactions to the increasing 'digitalization' of our work and our play is a leap back into the comforts and sociability of craft and artisanal approaches to among other things food and clothing. A resurrection of 'old skills' -- not only for the pleasure of doing things the 'slow way' but for the imperfections inherent in the objects created.
Imperfection is beautiful. Uniqueness as an artifact of the human hand may be a sentimental value for some -- but it is also linked to sustainability.
I'm a technologist, not at all anti- tech -- but I am also a humanist and prefer an artisanal loaf of sour dough!
I think the most important aspect of open source industrial design is the potential for collaboration between many fields beyond just design. For example, many objects are designed exclusively for other designers. Look in any design magazine, at any student or professional competition, and you'll see many objects that will never be mass manufactured or accepted and adopted by the public. Designers end up talking more about form, integration, and "honesty" of design more than durability, materials, and purpose.
But imagine an open online community where designers can apply their skills and resources to solving pressing social problems as defined by political economists, historians, artists, and scientists. Open source design has the potential to make design accessable to many people currently unskilled in design methods, and to bring design applications to areas that need it most.
RE: "hobbyist fabbers sharing their GPL'd designs and extensions around the world"
If Noble is on point (and I haven't read him yet, either) the historical importance of the machinist example may be that heavy pressure from economic and political forces (defending invested capital expenses; protecting industrial constituencies) may shape the way that instantiation hardware is R&D'd, as well as the way that "workers" (as opposed to 'hobbyists) interface with this movement.
In other words, over the next 10-15 years, if there are clear ways to block the production and widespread uptake of fabbing hardware for use by collectives and union-owned businsses (and I suspect that the usual suspects will find those ways), then the open source hardware movement will be overly limited to design wikis and the loyal hobbyists that roam there. OSH will be automated and most likely, centralized, minimizing it's potential to (a) create jobs; and (b) create user-designed goods that disrupt existing capital investments (cool stuff, like open-source thin-film solar or open-source batteries)
Hobbyists will have an important role to play, but if they don't interface with questions of job-creation, they will decrease their impact.
Now we're into an interesting discussion...
Re: Artisans and Machines
Maybe it's not that there's beauty in imperfection so much as inadequacy in mass produced desigs? The artisan will continue to improve on every subsequent piece that they create. Computerized assembly cannot. Computers handle data and information but knowledgeable people are always needed to turn those into useful work and decisions. Everywhere some form of thinking is needed, technology only ever augments what people can do. It does not replace. In the case of distributed/shared manufacturing, I would argue that tech is augmenting user's ability to meet their own material needs.
The fact that artisans and apprenticeships have fallen out of style seems to be a result of information-centric thinking insteaed of seeing the need for knowledge transfer. The envisioning of tech-centric, a-human utopias are another example of that. This thinking may be "western" cultural bias as well (see "iemoto" in http://itmatters.com.ph/columns.php?id=talisayon_091602)
Re: The "usual" obstructionist "suspects"
"if there are clear ways to block the production and widespread uptake of fabbing hardware ... the usual suspects will find those ways), then the open source hardware movement will be overly limited ..."
As an engineering design researcher, I rather disagree. If a design spec is robust and well described (not just a list of parts but a description of each sub-system's function, backround physics and design reasoning), I don't see how anyone could stop you from throwing together a fabricator from the huge surplus of technology that we're already drowning in. So you want to make a 3D fabricator? Start with an old photocopier:
- Lots of rollers and gears, ready made
- dozens of precise stepper-motors
- semi-fresh supply of high-purity carbon particulates (toner!)
- 1 hopefully not-burnt out high-voltage power supply for ion deposition.
Ok, so that's not a full design spec, but you get the idea. Once you have a few of these MacGyver'd replicators floating around at the local design co-op, it's just a quesiton of the microprocessor development model: make a basic one, then use it to create the next gen, then the next one...etc...
Further on artisans and machines:
There are still blacksmiths (Whom Noble's machinists replaced), carpenters without powered tools, people who bake and cook things from scratch, potters, weavers and many others. They are rare and do not comprise a significant portion of the economy. The things they make are much more expensive than mass produced goods. But I'd argue that our economy, having changed so much, actually makes it possible for these traditions to continue. In fact there are even academic projects now to rediscover and revive the techniques used by the ancient Greek shipbuilders and medievial cathedral builders.
I just had this crazy vision of academics, a hundred years from now, attemping to recreate a boiler or automotive factory circa 1920 just to rediscover things were done back then. Recreating and staffing a steel mill as a hobby? Astonishing.
Even in a world of garage and desktop minifactories, there will be people who will pay have a skilled artisan to hand blow glass for them. People will pay a premium for something that is made by hand. Yes, this is rather elitist but, I think it will sustain the art, the crafts and the techniques of ages gone by.
Maybe that's not what some want but I think that's the best we can hope for.
"if there are clear ways to block the production and widespread uptake of fabbing hardware ... the usual suspects will find those ways), then the open source hardware movement will be overly limited ..."
This takes us into DRM, which is really the only way they can do this. I think the current situation is instructive.
It is true that most people don't really care and will buy operating systems and applications (Microsoft Windows, Apple's iTunes, Adobe's document applications, etc.) whose defaults are set towards DRM. But this really only restricts piracy somewhat because there are knowledgable technicians out there who know how to bypass, avoid or subvert these all these measures. Current measures in scanners and printers make it hard but not impossible to counterfeit currency, you can rip DVDs in Linux with DeCSS, crack game software, etc. etc. etc.
The result is an endless arms race.
If the powers that be try to concentrate production and limit desktop fabs, they'll only succeed in creating a criminal economy. Or, to get very alarmist, they could create a police state where everyone is watched but no one is allowed to watch the watchers.
On the other hand, maybe that's all the powers that be want. They know that can't really stop piracy dead but they can at least make it rare enough that their bottom line isn't hurt that much--a unavoidable but minimizable expense of doing business. In such a world, it may be legal for a multinational corporation to reverse engineer something but it will be illegal for an individual to do it. There are criminals that do but they are hunted and kept rare. That seems to be status quo now.
Is this situation unavoidable? Can we replace it with something better? IP law makes my head spin because it seems to be ultimately arbitrary and bizarre.
RE: "This takes us into DRM"
I agree. Except DRM on steroids. Imagine the sort of (legislative / prosecutorial / patent law) full-court press that would be mounted to protect the market share of, say, energy producers or the auto industry -- industries that dwarf the RIAA by orders of magnitude. All it will take is a nidgy little NIST recommendation buried deep in a white paper saying that fabbers must adhere to certain standards and all pertinent patent law, and kapow! all of those cool old copiers hacked into fabbers are illegal.
Anyway...that's what I would do if I were a 'usual suspect' with $10b invested in the hardware sector and a whole stable of lawyers and lobbyists...
Perhaps this is why fabbers will take off in countries where such restrictions are less evolved. Viva globalization!
There is a determinism in the materials presented on this site and many like it that limit the discussion to prescriptive solutions when solutions is inherently wrong headed thinking. I dont mean to be critical because there are many useful ideas, some even highly creative, presented herein, but paradoxically those clever solutions merely perpetuate and undergird a system that is fundamentally flawed. Education is based on a stovepipe model that is very strong at building on the sequential accomplishments of the past. However it is weak on cross-pollination of ideas. The scientific network compensates for the lack of a systemic provision by webbing for this most essential aspect of a progressive and flexible system. It does not get it done. It is not surprising as it is based on the well known and observable evolutionary model used by biota of the planet. It is, however, inadequate for the situation that we find ourselves, i.e. humans facing probable step environmental change.
There are models in the universe that can be observed by humans by the manipulation of the rate of change. Most environmentalist have managed to assimilate the first small step of exponential rates and a few are conversant with the fundaments of chaos theory which is likely to be a useful tool for analysis of what has not even been named adequately, some call it surprise which at least hints at the unexpected as a common characteristic of some possible future.
In my opinion, one of the first things that needs to be done is to promote a change in the modalities that lead to inspiration, a revolutionary juxtaposition of human capabilities that exposes the full creative potential of the human consciousness. Martin E.P. Seligman promotes happiness and positive thinking which are necessary but not sufficient, although the concept of creating an openness to new modes of thinking is likely pointing in the correct direction. On the other hand, direction is a major problem as it easily becomes confounded with progress as a moving-forward concept. I am not talking about change, rather it is more closely described by the word transformation. Daniel Pink fumbles with right-brain, left-brain conceptual distributions. Beginnings, but basically object oriented as are the Western Languages such as English, thus it will be necessary to confound the confounding tools we so unconsciously culturally adopt before we can begin to problem solve thence find solutions. As a very basic example of what I am getting at I offer this poem about a well recognized poet that I know. Poetry in this poem is the metaphor for our accepted ways of thinking:
On hearing Kloefkorn
Former Nebraska Hog Calling Champ
Judging by the last part
of the last part
of his next to last live interview
I heard an old man talking
Boasting how strong and virile
Talking about the good times
his memoir records
how poetry just isnt sufficient
anymore, necessary but
Now that I am senile
how to misspell his name