Mon

Nov 20
2006

Tim O'Reilly

Tim O'Reilly

The Significance of Threadless.com

One of the recurring themes on the O'Reilly Radar is that of "news from the future," the idea that, as William Gibson put it, "The future is here. It's just not evenly distributed yet." We look for events and people that give us signals about what is to come. I've recently been thinking a lot about threadless.com in this vein, and working them into my talks about Web 2.0 as a stunning extension of the principle of "harnessing collective intelligence" into the design of physical goods.

If you're not familiar with it, threadless.com might be described as a digg for t-shirt designs. Users submit ideas for t-shirts, which other users vote up, till the most popular are manufactured and sold. The site would be cool enough just as a social networking site. There's a rich community, and everyone is having a blast. (Kathy Sierra ought to give it a shout out on Creating Passionate Users!) But think about it for a moment: This is a "crowd-sourced" manufacturing business.

Right now, threadless is just making t-shirts. But custom fabrication devices like laser-cutters, water-jets, and 3D printers are currently at about the price points of typesetting machines back when desktop publishing took off in the early 80's. Even traditional manufacturing techniques can now be harnessed by small companies and individuals, who can hire overseas factories to make short runs of custom designs. How far off is a future in which the creative economy overflows the thin boundary that separates "information" from "stuff"?

We've been fascinated with this idea since Marshall Burns and James Howison gave a talk entitled Napster Fabbing at our first P2P Conference in 2001. They pointed out, quite rightly, that in a world of personal fabrication machines, stuff could be shared as easily as music is shared today. But what would the mechanisms be by which new designs first come into play? Will they merely be copied from traditional manufacturing and brands, or will there be a new economy in which users compete in creative abandon? We've watched sites like shouldexist.org [which seems to be down] and kancept.com as lazyweb precursors to threadless. But to my knowledge, threadless is the first to have put all the pieces together.

It seems to me that this is one more step towards the reputation based economy that Cory Doctorow described in Down and Out in the Magic Kingdom. And of course, it's an important next step in the re-engagement between computer technology and the physical world that we've been so eagerly chronicling in Make.


tags:   | comments: 13   | Sphere It
submit:

 
Previous  |  Next

0 TrackBacks

TrackBack URL for this entry: http://blogs.oreilly.com/cgi-bin/mt/mt-t.cgi/5052

Comments: 13

  Michael Buckbee [11.20.06 08:55 AM]

You mentioned the hardware side of the equation, but equally important is the proliferation of accessible software applications like SketchUp and SecondLife that dramatically lower the barriers to 3d creation.

Also, checkout the work we're doing at Fabjectory - http://www.fabjectory.com.

  Dennis Linnell [11.20.06 09:04 AM]

If this "wisdom of crowds" approach sells t-shirts, why not use it to sell books?

  adamsj [11.20.06 09:27 AM]

Cory's recent story "Themepunks" (his bid to be the new Heinlein) is more on point for this. I gather it's a three-parter headed toward a novel (shades of fifties magazine publishing) I'm anxious to read.

  Frank Piller [11.20.06 10:39 AM]

Interesting post, however, there is a huge difference between Threadless and custom fabrication: The beauty of Threadless is that it is almost reversing the long tail trend (in production, not in creation). Threadless found a business model that allows companies to continue with good old mass production – not mass customization. A French company, look-zippy, has even perfected the Threadless model (more on http://tinyurl.com/zc76a.)

Another large t-shirt site, Spreadshirt, on the other site is doing this. They are like a personal fabricator for fashion. You create, and they manufacturer the stuff for you in the smallest (or largest) possible quantity. But they are lacking the collective filtering and peer-review of Threadless.

  jkottke [11.20.06 02:08 PM]

Tim, Skinny Corp (the proprietors of Threadless) also run Naked & Angry, where they print non-tshirt objects based on submissions. Right now, they've got neckties and wallpaper.

  Keith [11.20.06 06:22 PM]

Eeek! Don't feed the domain squatters. The URL is www.shouldexist.org.

  Tim O'Reilly [11.20.06 07:32 PM]

Keith -- that's what I remembered too. But shouldexist.org gives a connection refused, while shouldexist.com is in fact a listing of "should exist" kinds of things.

  Susan Wu [11.21.06 12:54 AM]

Artur Bergman and I were chatting about this the other day - that essentially what happened in Second Life with CopyBot is one day going to happen in the physical world. We will have 3D printers and scanners who can scan molecular architecture, thereby revealing the source code of these objects. Objects and content will be commodities, but the scripts (the human experience) will never be.

  Tim O'Reilly [11.21.06 11:34 AM]

Keith Hopper sent me the following notes by email re. shouldexist.org vs. shouldexist.com (and I've changed the original story accordingly):

"Looks like shouldexist.org is just down (perhaps a casualty). Shouldexist.com is definitely a domain squatter and the reason they pick up relevant links is because they use info from their (accidental) referrer links to auto-generate seemingly innocuous 'navigation'. It's a pretty deceptive little trick.

Here's some similar sites not mentioned in your post (lazyweb to the rescue):
globalideasbank.org
halfbakery.com
thinkcycle.org
yet2.com
www.idea-a-day.com

I've also been theorizing around distributed innovation alongside participatory technology in general at my blog:
www.hopperanalytical.com/blog"

Nice. FWIW, thinkcycle.org was started by our friends at Squid Labs back when they were still at the MIT Media Lab.

  Andy [11.21.06 06:21 PM]

Not that it's too important to this article, but Threadless has recently begun producing sleeved versions (hoodies and long-sleeved tees) of some of their designs.

  csven [11.26.06 08:30 AM]

While I find what Threadless has done interesting, it's a long haul from 2D graphics to articulating 3D forms. And while my own blog is very much about the collision between real and virtual product design and development (due in large part to RM), I'm not so sure there aren't some potentially difficult speedbumps along the way. Here are a few to consider:

1) when people vote for designs on Threadless they have a very good set of expectations for what the final product is, whereas an image of a 3D form tells a person only a relatively little bit about the design. A graphic for a t-shirt is effectively replicated on one's monitor. While I expect we'll eventually have display and tactile systems which allow average users to evaluate intangible three-dimensional forms, that may take a while yet. As an example, someone might ask: does that rubberized part feel good or does it feel clammy and disgusting? There are many things that I, as a product designer, consider that most consumers don't consider... and that's the goal: for them to not notice.

2) Threadless designers don't concern themselves with material, yet even in established manufacturing processes material decisions can be complex; there are a quite a few different major kinds of plastic for example (LDPE, PP, SAN, ABS, etc), and a great many blends of each of those... and for each blend there are differently engineered characteristics. Rapid fabrication processes are still developing materials and have a ways to go yet.

3) No one thinks about product liability with a t-shirt graphic, yet ask toy designers about that topic. Ask step stool manufacturers about it. Ask manufacturers of objects that interact with things we ingest about the certification process (and the cost). This could be a big issue.

4) Degree of difficulty in creating real objects versus a 2D graphic may short circuit a Threadless-style plan. Let's face it, most anyone can - with little effort - create quite a number of products already: silkscreened clothing, greeting cards, jewelry, Christmas ornaments, aso. Yet most people don't create things. In fact there's this "awww, look" response at gift-giving gatherings when people go through the trouble to create something (and sometimes - perhaps quite often - the effort is unappreciated). How many people don't make their own holiday cards because either they lack the time, the skill, or the self-confidence? Many, I'd venture. And now we're expecting them to become product engineers and sculptors? I have my doubts. Having watched the videogame mod community take a severe hit with the release of increasingly sophisticated games contained complex 3D models, I see no reason to believe a "crowd" will be rushing to design their own 3D products any time soon. We live in a convenience culture, and at a certain point this activity becomes very inconvenient.

5) Warranty. When someone designs a simple shelving system that, while sitting in someone's Arizona garage in the middle of summer, warps and sags, the consumer is going to be unhappy. What recourse do they have? Depending on conditions of sale, they may not have much. This of course moves into issues of Reputation. Only right now we don't have a standardized system, and we don't seem to be having much luck creating a system that can't be gamed.

There are probably more, but that points out a few at least.

Now, on Napster Fabbing and in regards to Susan's comment, the issue of 3D data theft has been on my mind for quite some time (since almost immediately after the appearance of the mpeg layer 3 format). It's also the reason I did the videostream rip last year (that inspired the development of OGLE); it was a warning to other product designers. This is a real problem not just because of the reality of what it represents, but in the perception. Designers will be (and some of us have been) looking for business plans that protect our time and effort given that we can no longer reasonably expect IP laws and the legal system to help us. So far the options I've found aren't particularly appealing to me.

Finally,

"Will they merely be copied from traditional manufacturing and brands, or will there be a new economy in which users compete in creative abandon?"

This one is easy to answer based both on observation and on the above points: people will copy. It's easier and it's what they already do. Just spend some time in the 3D community and see what they're modeling. Plenty of unique character designs, but very little product originality.

  Alexzander Macpherson [06.22.07 01:50 AM]

Actor Bruce Willis says he was knocked out while filming Die Hard 4.0, as he attends the London premiere...

Post A Comment:

 (please be patient, comments may take awhile to post)






Type the characters you see in the picture above.

RECOMMENDED FOR YOU

RECENT COMMENTS