I'd like to start with a quote by the Taiwanese author Wu Ming-Yi: Her voice infused people with the essence of a song. It would turn into a windblown seed: you never knew in your heart where the seed would fall, nor when it would sprout. This may seem like an odd quotation for an industry conference about location tracking, pervasive identity and generally solving for who's on first. I like this quote because it nicely describes some of what — and why — we are trying to do with the Pen.

I grew up in Canada.

Canada does not have the kind of creation myth, or foundational purpose for being, that a country like the United States does. No one has trumped the US when it comes to creation stories but Canada was birthed, at nearly the same time, entirely in the service of economics and geopolitics. Canada was created first in the service of the merchant classes and second to ensure that a British railroad reached the West Coast before the Americans thus preventing them from going North. Or maybe the other way around, but you get the idea.

Despite all the Canadian Heritage Minutes that have been produced in the intervening years it's not really an inspiring start. What is interesting about Canada though is that it's managed to do pretty well despite its generally crass beginnings. Canada is not without problems but, at least in modern times, has managed to create a social body that mostly works most of the time for most of the people.

It sometimes seems that the most critical thing people can say about this arrangement is that it is boring but we'll save a discussion about people being adrenaline junkies for another day.

It is important to mention that in Canada most anything has historically mostly always been to the exclusion of the First Nations for mostly everything. Canada's relationship with the First Nations, until recent years (and still now in some cases) can only be described as appalling and shameful. Since I started writing this blog post the Truth and Reconciliation Commission of Canada mandated to tell Canadians what happened ... and to create a permanent record of what happened in the Indian Residential Schools has been published. I have not read the findings yet but I will and so should you.

Note: This is the official Smithsonian logo but just in case there is any confusion this is not the official Smithsonian motto. Anecdotally, though, this would be the most popular staff t-shirt ever printed.

I started with a story about Canada because I see parallels in the origins of the Smithsonian. The Smithsonian was not an American creation but rather the unwanted dream of an English businessman named James Smithson. Sasha Archibald's The Difficult Bequest: A History of the Smithsonian" in the LA Review of Books sums it up nicely:

The Americans didn’t ask for Smithson’s charity, and neither were they glad to receive it. Congress had more pride than greed, and the unexpected gift rankled: not only was it that of a reviled Brit, but a Brit who dared demand he be acknowledged in perpetuity. Moreover, it was earmarked for a purpose Americans never would have chosen themselves. Smithson’s patronage was condescending — nothing more, one Congressman surmised, than a rich man’s bid for immortality. Even John Quincy Adams, the bequest’s most passionate advocate, refused to venerate Smithson as a magnanimous patron. It was Adams who kicked up a fuss when investors were allowed to squander the funds (later replenished by the US Treasury) and Adams who protested that a national farm didn’t meet Smithson’s stipulations. In private, however, he concurred that James Smithson was probably insane.

But you know 169 years later the Smithsonian has evolved in to something pretty awesome. We have all the stuff or, more precisely, 137 million objects and counting. The Smithsonian is a public instution with a mandate to act as a public good and there aren't many other cultural heritage organizations that operate at the scale or with the breadth of the Smithsonian.

I also like to joke that the Smithsonian has the world's largest conceptual art collection. The good news is that there really are 137 million objects and they are actively being cared for. The bad news is the only real proof of this that you, as a citizen, have is a leap of faith when I, as an employee of the Smithsonian, tell you these things are true.

That's the larger context that the work we've been doing around the Cooper Hewitt's much smaller collection website operates in. We have 217, 000 objects and have been using the collection website, and the idea the every thing of importance to us has a stable and permanent URL, as a device for people to have common referent — something they can share — for the real-life objects they may not ever be able to see.

The collection website is also the scaffolding on top of which all the work that we've done for the museum's re-opening sits. If the macro-question is how do you provide access to the entirety of a collection the micro-question is: How do you provide recall for those parts of a collection that are actually on display?

What if you could come to the museum and remember your visit without having to spend half your time futzing around with devices to cobble together a half-assed and short-lived memory?

Think: Taking pictures of wall labels or taking pictures of objects without necessarily knowing what they are or who did them or how you figure those things out once you've left the building.

Which is why we built the Pen.

I'd like to start with an abbreviated timeline of the Pen project. In September of 2012 we launched a new version of the Cooper Hewitt collections website. In March of 2015 we launched the Pen, an NFC-enabled capacitive stylus given to every visitor to the museum.

In about Decemer of 2012 we launched a public API for the collections website. It didn't actually take us three months to build the API. It took us three months to decide that it should be at the top of the TO-DO list.

At the beginning of February 2014 there was, for all intents and purposes, no pen. If you had asked us to demonstrate the Pen we would have only been able to show a whiteboard marker with an RFID tag taped to it and a Raspberry Pi with an RFID antenna and a single hard-coded object. There were other efforts in play that ultimately came to pass but at the beginning of the month there wasn't much tangible to demonstrate.

This is a exagerated but not inaccurate description of what all of 2013 looked like.

By the end of February 2014 we would have not one but two different functional prototypes for the Pen and less than a year later we managed to design, produce and manufacture custom electronic hardware from scratch. (The finished pens arrived at the museum a few weeks before they went out on the floor.)

There are two lessons to take away from this.

The first lesson is: It turns out that it's possible to do this kind of thing in 2015. It turns out you can do capital-P captial-D product design and manufacture something in a meaningful volume — say more than 20 or even 100 and less than 500, 000 — for a reasonable cost.

It's still really hard, requires a stupid amount of work, no small amount of shouting at people and a not insignificant capital investment. It was more than lunch money but it was not an irresponsible amount of money and if you imagine that the work we've done might some day be deployed to every other Smithsonian museum on the Mall then the costs start to seem laughable.

There is no plan to deploy the Pen on the Mall yet but we should. The implementation details are non-trivial but it's totally do-able (really) and would be a good way for the Smithsonian to exercise and exorcise some of its institutional demons and generally just be awesome for everyone else.

The second lesson is: Don't do what we did. Don't try to do this in a year. It's possible but there is no valour in the burden we took on. It's a stupid way to work but we (the museum) had painted ourselves in to a corner from which the only escape was to walk back out again, painting over our footsteps as we went.

I choose to see our decision to suffer that indignity as proof that the Pen was worth doing but that's also the point: The Pen was worth doing on first principles not because we have War Stories to tell after the fact.

The reason I mention all of this is that the constant throughout this entire process was the collections website and the API which are really the same thing. I sometimes hear people saying that APIs are hard and I am pretty sure I don't agree with that. We launched our API three months after the collections website and a full two-years befor the musuem re-opened.

This is important. An API is not a finished anything. That's the point. An API is access to data in the service of an idea that does something. I think there is a popular myth that when an organization publishes an API it will somehow magically self-assemble in to a finished product adored and embraced by the masses.

I think that a more useful way to understand APIs is that they exist to allow other people to build tangible proofs without needing to badger you all the time for the details.

This is not about the client-services model itself. There is nothing wrong with the client-services model and lots of good work has been produced under its umbrella. It is about the roles and responsibilities, and their consequences, that museums have forged with client-services companies.

It is about the reality that museums don't know how to do the things they want to do and are entirely dependent on client-services companies. It is about the reality that museums don't even know how to talk practically about the things they want to do outside of magical unicorn deliverables or wishful thinking or simply burning a problem away with unsustainable amounts of money.

It is about the reality that those same client-services companies can be forgiven for not believing anything their clients say (goals, funding, tolerance for risk) and for ruthlessly pushing the work through a meat-grinder tailored to their own schedules and capacities. That's just business.

It is about the reality that if any one of those dynamics changes there are outsized ripple effects on all the remaining dynamics. If we're going to assign blame or responsibility though I think we would do well to start with ourselves: The sector has allowed this to happen to itself.

For those of you who don't already know in a past life I worked at Flickr for about a million years. This is a slide I made to describe to way I understood things to work at Flickr, circa 2009. One of the disconnects between the way we worked at the way a lot of other parts of Yahoo (who acquired Flickr in 2005) worked was this idea of products versus features. Most of Yahoo was instrumented to treat features as stand-alone products and not just an on-going extension or faceting of the original product, in our case Flickr.

This is what most feature launches looked like. If you were an engineer the two-to-four weeks prior to launch were Pure Suck because everything kept changing with less and less time to actually implement those changes. Almost every part of the engineering culture and infrastructure was designed to accomodate this reality and it was a big part of our success but, in the moment, it totally sucked.

This slide is incomplete, though.

See the way the line keeps moving forward after launch? This is where the Burden of Suck was transferred from engineers to product managers. This is part where the thing that launched, after all that hard work, keeps changing. And then it just gets rolled back in to, and affected by, the larger project itself.

So I think the lesson in all of this is as much about lowering the cost of failure as it is lowering the cost for success. The Pen was not as many, including the museum, believed simply a question of snapping together a suite of off-the-shelf hardware and software pieces like a kind of cultural heritage Lego model. It was a lot of trial and error shoehorned in to a brutal schedule which meant a ridiculous amount of energy and attention was spent worrying about getting something right the first (and only) time. It's pretty great that we pulled it off but the number of places the whole project could have gone completely wrong and failed were so numerous it is sort of a wonder that we chose to do it at all.

Hail Mary passes are exciting but they are a bad way to run a project.

The reason I started with a timeline of the API is that in building out our own infrastructure, fast and early, and likewise maintaining it ourselves we have had a practical way to build working prototypes as a means to address unknowns and, more importantly, debate. Put another way, we had a way to make the time and the effort to disprove an idea or a concern or an outright argument so low that any questions about testing it in the first place were rendered moot.

Another way to see any discussion about the cost of failure or success is to understand it as measuring the cost of liability and the practice of outsourcing risk by delegating all responsibility for anything that looks or feels like a failture to third-party contractors. As a practical matter though the cost of minimizing the wait-time between a concept and something that works is very different for an organization than it is for a contractor.

Those costs become even more pronounced when you are dealing with something as complex, and as novel, as the Pen was and continues to be. The following passage is from the paper Seb and I wrote for MW and tries to address the issue of those costs:

It is important to recognize the problems with this model (a one-off innovation to a commodity – ideally one which can be bought cheaper from a third party supplier and supplimented with a support contract that outsources liability and risk) when it comes to services-based projects like The Pen and its ‘museum-wide’ system:

• Support contracts are typically for projects whose value diminishes over time (because on-going development of a project stops at launch) and so the cost of that contract, in effect, increases year over year with fewer and fewer returns.

• With support contracts very often no institutional knowledge about the project or how it might serve as a platform for future projects is gained. Sometimes the only tangible skill an institution learns, or passes on, is how to manage contracts for service providers.

• A support contract is only valid for past work. Any additional changes or improvements incur additional non-trivial costs because third-party contractors not only need to complete the work but require time to familiarize themselves with a pre-existing code base.

• Support contracts represent money that leaves the institution and never comes back.

These overall costs are likely to be much more than it would cost to hire and maintain permanent creative technical staff in a museum when intangibles like institutional knowledge and the ability to build on the value – technologies, learnings, especially created by past work – are factored in.

As a sector we have spent a couple of decades making excuses for why “digital” can’t be made core to staffing requirements and the results have ranged from unsatisfying to dismal.

The shift to a ‘post-digital’ museum where “digital [is] being naturalized within museums’ visions and articulations of themselves” (Parry, 2013) will require a significant realignment of priorities and an investment in people. The museum sector is not alone in this – private media organisations and tech companies face exactly the same challenge. Despite ‘digital people’ and ‘engineers’ being in high demand, they should not be considered an ‘overpriced indulgence’ but rather as an integral part of the already multidisciplinary teams required to run a museum, or any other cultural institution.

The flow of digital talent from private companies to new types of public service organizations such as the Government Digital Service (UK), 18F (inside GSA) and US Digital Service, proves that there are ways, beyond salaries, to attract and retain the specialist staff required to build the types of products and services required to transform museums. In fact, we argue that museums (and other cultural institutions) offer significant intrinsic benefits and social capital that are natural talent attractors that other types of non-profits and public sector agencies lack. The barriers to changing the museum workforce in this way are not primarily financial but internal, structural and kept in place by a strong institutional inertia.

Good times.

Years ago, I read an interview with the director Hal Hartley. If you've never seen any of his films they are punctuated by a noticeably weird and unnatural style of speaking employed by all the actors.

Hartley talked about how, following an argument, many people spend days and days, thinking of the perfect come-back or rejoinder to something that was said to them. I try to write my dialog, he said as if each character had that perfect come-back on the tip of their tongue everytime they spoke.

One of the things I've been trying wrap my arms around is the belief, in almost all large organizations, that a project once launched will never be revisited. This is especially true in politics or at least perceived to be true, which might be worse.

Despite all the arguments someone else might make in its favour there persists an institutional cynicism that version two of a project will never happen.

As a consequence everyone tries to shoehorn their part of a project in to version one with all the corresponding complexity and bloat and politics you might imagine that entailing.

You can see a parallel to these phenomenom in the way that museums think about visitation and, in turn, visitors think about museums. In short: The belief that everything needs to be crammed in to a too-short 90-minute visit and it all needs to be meticulously documented as a consequence since the whole experience ends up being like a cultural heritage version of shock and awe.

There is a spectrum (of both museums and visitors) of course but what underlines both activities is a belief that the museum and the museum visit will never be returned to in any capacity once someone walks out the door.

There is ample evidence to back up this observation but what if you could come to the museum and then leave with a record of your visit that didn't demand you spend half your time futzing around with your phone or some other device?

How would that change the ways that you observed and interacted with the objects on display and how would that changes the ways that the museum interacted with you?

What if you could go to the museum and have the confidence of memory or recall? What does it mean for a museum to say to its visitors We'll keep a record of your visit and make it available for you to re-visit at some later date? We won't preserve it the way that we do the objects that have been accessioned in to our collection but otherwise it's no trouble.

Maybe that is what part of what being a museum in 2015 means. Certainly for a public institution like the Smithsonian I think there is reason to think that's what it should mean. Our entire purpose is to keep things safe, after all. What if we extended that courtesy to your visit?

So that's why we made the Pen.

Which is in many ways just a very elaborate physical bookmarking system. People, it turns out, like to bookmark things.

In a world where everything is measured against the benchmark of Facebook's or Google's scale and success our numbers are pretty humble. On the other hand if you compare the those numbers to the numbers we used to enjoy — attendance, dwell-time, repeat visitation or any indication that a visitor didn't immediately turn to dust the moment they walked out the door — it's been a raging success.

The figure that I often mention is that, in the less than three months since the Pen has been available to visitors, there have been just shy of one million act of collecting. Which is pretty cool considering that there are only about 700 objects on display in the museum itself.

And of those million acts of collecting visitors have collected 3, 900 different objects. When you look at the numbers it is absolutely clear that most of the things that people collect are collected from the wall labels, rather than the interactive tables, in the galleries. I can't be bothered to do the math to establish whether it's really an orders of magnitude difference simply because the numbers are so extreme.

But think about it: Most things are collected from wall labels but there are only 700 objects on display but 3, 900 distinct objects have been collected.

Part of the workflow for any new exhibition involves the curators choosing, for every object on display whether its a loan object or one of own, up to ten related objects from the collection. What related means is left up to the curators and the only criteria is that an object on the floor exists within a larger context or institutional velocity.

In addition the curators were asked to tag each one of those objects (on display and related) using a taxonomy of their own choosing. And this is not meant to be a one-off either. We do this now for every new exhibition going forward which means we are always supplementing and improving past work.

That people are collecting 3, 200 objects not even on display in the galleries points to the larger opportunity in doing something like the Pen.

Every act of collecting an object (or of creating your own on one of the interactive tables) has a permanent and stable URL. What that offers both the museum and its visitors is a tangible and practical scaffolding that affords each of those permalinks the same possibility to serve as a resource the way the interactive tables do in the galleries.

This is not work we've done yet nor it necessarily obvious, easy or inevitable. We have been a little busy just making the basics work but we have always done that work with an eye towards what it makes possible next.

During the question and answer for this talk someone asked me what we thought our visitors would do with all of the stuff they collected. My answer to that question has always been: We have confidence that our visitors will figure out what to do with it on their own terms and for their own reasons.

The list of things that you might hang off any one permalink for an object collected is mostly limited only by the time and typing required to implement it so I don't get hung up on any one specific thing. We should do all of it.

The bedrock of cultural heritage is debate and nuance. Neither of those things square well with the demands of architecture nowhere better seen than in the practice of asking people with decades worth of theory and practice to reduce their knowledge to a scant paragraph taped to a wall next to an object.

Rather than trying to usher in a paradigm shift in the realities of physical space the Pen, and all those permalinks, offers a middle ground. There is still a primacy in the object itself (even if that object is a device for a larger social gathering or event but now it shares that space with recall) with the safety of staying present in the moment confident that it might be revisited later, with the luxury and the convenience of being able to choose what later means.

Maybe later also includes the ability for museum visitors to contribute to the meaning of an object or an event. As a design museum we increasingly traffic in things that are literally hollow shells without the software or the data that runs them. Privacy details aside, what if you could donate the data your Nest thermostat collected to sit alongside the same hunk of metal and plastic (and concept sketches) that we have sitting in our collection? What if there was a way for you to participate in the meaning of that object?

Which is all fine and good for museum objects, right? I like to think so. This is what the fruits of 2015, all of this amazing technological capacity to suspend time, make possible. The batteries not included part of that equation though is the social contract around that kind of recall.

Pretty much everything I've just described, done to foster learning and understanding in and of the arts, is also being done by big companies and intelligence agencies alike. If Edward Snowden is to be believed then it is being done by the latter at a scale and with a reach that sometimes exceeds imagination.

Let's assume for the sake of argument that the stories told about the NSA's ability to record and recall history are true. Let's assume that their ability to weave tapestries of meaning out of nuance and innuendo is unparalleled. When all of that attention is directed at any one individual they might be forgiven for thinking the whole project is a little sinister.

There's the rub, though.

When you remove the sting of present consequences the ability of these organizations to record your visit so to speak starts to look not just attractive but beneficial in the long run. In some distant future where our present is just a shadow of its current concerns they will thank us for our ability to preserve ancillary materials associated with a subject and of instrumenting the cost of following hunches and generally spelunking through the mass of data to the point of seeming negligible.

These are not the issues that governed the Pen, or the museum's re-opening, but they are the places where the Pen overlaps with the present. These are not tomorrow's problem.

Discuss.