Hi, my name is Aaron.

Thank you for inviting me to speak. Because it's just me up here, alone on the stage, speaking to you I want to take a moment to acknowledge that I am only one person who is part of a larger group working on all the things that I am going to talk to you about today.

The larger group is the museum itself (and a little more abstractly the Smithsonian) but day-to-day that group is the Digital and Emerging Media department where I work. We are coming out of a pretty intense three years of work and just as importantly a thinking about that work and there is no way that I am going to do all of that work justice in a one-hour talk.

We have tried to be as open as possible about the work we've been doing, including the false starts, and have made an effort to write it down on the Cooper Hewitt Labs blog. When you have a chance I would encourage you to spend some time going through the archives. There's a lot there, written by a lot of different people, and hopefully it will add even more texture to everything I am going to talk about now.

I arrived at the museum in July of 2012. I moved from San Francisco to New York City for three reasons.

First, it was and still is a good time to be a design museum. Second, it was a chance to work with Seb Chan. Third, it was a chance to work with Seb Chan on a design museum that was asking itself what it meant for a museum to be part of the internet and vice versa with Bill Moggridge.

At the time I remember reading a profile piece about Bill in which he said I don’t know about museums, but at least I know design which, for me, became the fourth reason to sign up for the project: What would it mean for a museum to be part of the internet in the service of its subject matter rather than the other way around?

Bill was forced to take a leave of absence, for medical reasons, around the time that I started at the museum and he died a few months later. It is I think — I hope — a testament to how much the everyone valued and respected Bill that we tried to take on the project he had charted for museum without necessarily knowing what, exactly, he had in mind. I don't imagine that Bill, himself, knew at the time.

We have accomplished a crazy amount of work in the last three years and we have birthed a thing that will require probably as much or more work going forward to ensure it is not just a one-trick pony. I am going to talk to you about some of that work but first I am going to take you through a bit of a wandering preamble. I want to try and show you that this is not work that was done for the sake of the work itself but in the service of the museum and, in particular, a design museum.

Before I tell you what we did and a little bit about how we did it I want to explain why we did it, or at least why I think we did it. This was not an easy project for any part of the museum and there were lots of moments when we could have decided that it was too ambitious a project given all the other constraints but we didn't and I would like to believe that we didn't for good reason.

Also, that's not a picture of Bill. It looks a bit like Bill (and that is something that Bill was famous for saying) but I am not going to tell you who it is because I think it will be funnier when you figure it out on your own...

I'd like to start with three quotes, none of which are immediately related to one another. They all cast a shadow on one another, though, so hold on to them throughout the talk like a kind of soundtrack music.

The first quote is by Aaron Keefer, the gardner at the French Laundry restaurant in Napa. He said: I think in any cooking, what we really do is move water molecules from one place to another.

The second quote is by Ali Soufan, a former FBI agent: Imagine if we didn’t go down that road. Imagine. We played into the enemy’s hand. Now we have American hostages in orange jumpsuits because we we put people in orange jumpsuits.

The third quote is by @selfawareroomba. We actually have a Roomba in our collection and we were eventually able to convince it to write an object of the day blog post about itself: A ROOMBA sees the world in math / And tries to rid you of you / And the trail you left behind

Lately I've been playing with the idea that history itself is the space left over as any two moments in time tear away from each other. Or as they fade the way a mural in the sun gradually disappears; people both aware of its disappearance and shocked when it finally vanishes.

There are still stand-out events (the clues) and we recognize those in the objects and artifacts we celebrate. More specifically that we celebrate those objects in common. The scarcities of the past meant that the pool of common celebrations from which to choose was pretty limited and but even now, while it might seem like we're swimming in tailor-made niche rituals, I don't actually think the fundamental dynamic has changed.

There is still what Scott McCloud dubbed the magic in the gutter. The "gutter" being the space between any two panels (or frames) in a comic strip. The gutter is the place where the author and the artist let the reader act as the narrative bridge between two events. This is an integral part of comics as a form and I think fundamental to their popularity.

I like to think of the gutter as the space where fan-fiction operates. As a way of creating alternative reasons to explain why any two events are related to one another.

I've also been thinking about something I'm calling Interpretation Roomba. This stems from an on-going discussion I've been having, mostly with myself, about the 9/11 Memorial. The 9/11 Memorial is a complicated space for many reasons: The question of whether it's a memorial or a mausoleum or a museum; The question of abstraction and signifiers in all memorial constructions; The questions raised by the event itself.

What I find problematic about the 9/11 Memorial is that I left feeling as though they have paved over all the spaces between the common or shared "peaks" that we use to recognize and remember the day itself. Paved over and inhabited by a kind of interpretive robot vacuum cleaner set loose to hoover up anything outside of a single master narrative. As if a space like 9/11 shouldn't cause you, maybe even force you, to leave with more questions than you entered with. I think the 9/11 Memorial is meant to be cathartic; it aims to evoke a very deliberate emotional reaction.

And manufactured emotion is a like a bad habit for a lot of museums.

Monkey Jesus. Let me start by saying: I love Monkey Jesus.

Monkey Jesus is sometimes known as Ecce Homo, a church fresco painted by Elias Garcia Martinez in the 1930s in Northern Spain. Like many churches in Europe it was abandoned and stood waiting to be reclaimed by the elements. In August of 2012 Cecilia Giminez, a nearby resident, decided that she would attempt to restore the painting before it was completely lost.

That last fact is really important: This fresco was the proverbial tree falling in the forest with no one around to hear it. Had another year or another decade passed the painting would have been washed away by the rain or the sun and no one would have known the difference.

What happened instead is that someone posted a picture of Giminez's efforts to the internet and the whole world when completely nuts. This, we were told, was an offense against all culture. Proof that the laity shouldn't be trusted with the arts. That this 90-year woman had single-handedly destroyed everything sacred about the Rennaissance.

Then a funny thing happened: By the end of 2013 forty thousand people had visited the church to see the fresco and Giminez herself was pushing for financial compensation claiming artist's rights for her work.

Monkey Jesus has crossed the event horizon of signifiers and now, I'm willing to bet, we're going to actively preserve the so-called failed restoration over the original fresco precisely because of this history. Because now the work has narrative pedigree. If you think that sounds like crazy-talk consider the exact same painting but done in the hand of Alex Katz or re-created by Cindy Sherman. Look carefully at Monkey Jesus and tell me you can't see the shadow of either artist's work in that painting.

The issue is not whether a different artist would have done Monkey Jesus better but in how we reconstruct the narrative around an event; the reasons we choose to understand why an object is worthy of a narrative at all.

I have been known to stand in front of audiences and announce that the Smithsonian has acquired the Global War on Terror (GWOT) as a design object.

This has always been a deliberate provocation and a set-up for the larger question about what it means for any museum to collect "digital" objects and in particular a design museum to collect things like service design or experience design. It's been a way to force ourselves to admit that we don't know how to collect things which are categorically real but which are not made manifest in any singular physical form.

How does one exhibit the GWOT, anyway? What are thoughtful and meaningful museum experiences for a subject like this? Do we have a room where you can be threatened by angry German Shepherd? Would it better or less awkward if it were an animatronic dog? How about an interactive waterboarding experience? I recognize that this all sounds like the theater of the absurd right now, still so close to the events. On the other hand I think we all know that someone will eventually collect the GWOT and then what? Could you buy an orange jumpsuit at the museum shop?

So that was the gag for a while but then in January of 2014 something a bit crazy happened. On the day marking the 10 years since the opening of the prison at Guantanamo a group of protestors entered the Smithsonian's Museum of American History and stood at the entrance to an exhibition about American service men and women, titled The Price of Freedom.

They stood at the entrance to the exhibition dressed in orange jumpsuits, shackled, wearing black hoods covering their heads. And when I saw the photos of the event without knowing the backstory I could easily imagine thinking that this was the work of a particularly brave or at least provocative curator.

Then, about six months later, the beheadings started happening. Not just beheadings but Americans (and others) being paraded around deliberately in the trappings of their own war efforts now being thrown back at them. The orange jumpsuit has become a design object. It has become a media device whose symbology is manipulated in the service of an intent, of a goal, of a design brief.

My speculation about the GWOT was always a rhetorical device but now, I think it's fair to say, somebody really does have to acquire the damn jumpsuit.

Who would acquire a thing like the orange jumpsuit? Normally it might be the National Museum of American History. They get all the cool stuff. More recently the Victoria and Albert's Rapid Response Collecting group might take up the challenge. I will come back to them later in this talk. But would they acquire two jumpsuits – one used by each team so to speak – or just one of them?

How much do we even know about the orange jumpsuits used in the beheading videos anyway? Are they just dollar store knock-offs or are they authentic US military prison issued jumpsuits stolen during a raid on one overseas prison or another?

Is this question important? Is it important in the way that the National Archives has Jacqueline Onassis' "pink Chanel suit"? You know, the one she was wearing the day her husband was shot and killed as she sat next to him in a moving vehicle. Maybe, by virtue of the magnitude of that day's events, there is forever now one only "pink Chanel suit" but that then forces the question: Is it important – as a policy decision – that the dress be kept out of public view for the next hundred years? Should the orange jumpsuit face a similar prohibition?

We haven't (yet) made the same determination of Trayvon Martin's hoodie. For a moment anyway it seemed as though it was set to go on display as part of the opening exhibits at the Smithsonian's National Museum of African American History and Culture. Said Lonnie Bunch, the museum's director:

It became the symbolic way to talk about the Trayvon Martin case. It’s rare that you get one artifact that really becomes the symbol. Because it’s such a symbol, it would allow you to talk about race in the age of Obama.

The thing about Trayvon Martin's hoodie though is that the whole idea of the hoodie itself has become a not-so veiled coded language (a referent) for a whole set of social and cultural prejudices that ended with a black teenager being killed for how he dressed. He was killed because of someone else's construction of understanding around an object.

The friendlier version of acquiring the War on Terror is the idea that we might instead acquire Virgin America. If anyone has mastered and advanced the craft of an all-encompassing service design in modern times it is them.

But again, what exactly do you acquire or rather how do you frame an acquisition in such a way that it might convey everything that VA was not ? Do you need to? How do you provide the context to demonstate what made their work significant? Is the inability to convey that context reason not to acquire something?

Like the War on Terror, acquiring an airline is a bit of a conceptual device and so I've often joked that what we might do is acquire the rights to fly the JFK-SFO route in collaboration with the National Air and Space Museum (NASM). The Cooper Hewitt should not be allowed to do this but NASM know a little bit about airplanes.

It's an intriguing idea. Setting aside some very real privacy questions for the sake of argument imagine all the metadata around a real-live operational airline that a museum might collect. Imagine what it might mean for a museum to take the measure of an acquisition in the moment. What if the object is its own curatorial file? What are seat-back pockets if not a place for an airplane to tell you about itself?

So one day while I was in Washington I tried to pitch the idea. It didn't go very far and in the process I was reminded of three things and learned another:

1. Museums really don't know how to run services. We should and we need to but institutionally we are still geared and tempered for preserving things in amber. We're really good at taking care of things that don't move.
2. OMGWTF liability. There is no way that our lawyers would let us do this which is probably a good thing. On the other hand it doesn't seem too far-fetched to imagine a scenario where Airbus gifts Virgin America a plane, as a tax write-off or something, and Virgin agrees to undertake the day-to-day maintenance and operations with the Smithsonian treating the plane as a great big living gallery. Eventually when the plane is no longer fit to fly it could be moved out to the Udvar-Hazy Center like we do with most planes.
3. We actually did this once. Back when the Concorde still graced the air we had an agreement with Air France to take possession of one their supersonic planes when they stopped flying it. The story goes that the plane had a little plaque in the cabin saying that the plane would eventually become part of the collection of NASM when it was retired. A wall label, of sorts.
4. Southwest Airlines.

This was the thing that I needed to be reminded of: Virgin America and Southwest Airlines are basically the same beast. One has been more successful than the other but ultimately they share nearly identical approaches to operations and an almost naked appeal to consumer archetypes.

So there's an interesting question: Is an awareness of the archetypes that guide design objects a precondition for design? Is part of what we love about design an appreciation of the craft of manipulating those archetypes in the service of spectacle?

Is a successful design object the one that succeeds in normalizing archetypes? The other way to think about that is to ask the question: Is a design object conferred notability because it is the first to pioneer a concept or an approach or because it was the first to succeed in the marketplace?

I think this is a relevant question for design because the issue of function and audience and of servicing a need that can be articulated are pretty central to the way that design distinguishes itself from other disciplines.

I think most of the interesting work in museums these days is happening under the umbrella of the V&A's Rapid Response Collecting project. I don't always hold the same regard for some of the things they collect but like Cecilia Giminez and Monkey Jesus they default to action which is better than the alternative.

One of the items they've collected is a pair of Primark cargo shorts made at the Bangladeshi sweatshop that collapsed in April of 2013. I pity the conservationists of the future who are going to have to preserve this thing. Unless they are the fashion equivalent of a Twinkie or a plastic bag and can sit unmolested on a shelf for a hundred years I imagine they will be an constant and on-going challenge to preserve. But that's okay. This is what museums do. We're good at it.

But how much of the context will we preserve along with it? In fifty years time, why will this pair of pants from the year 2013 be any different from any of the pants in the Cooper Hewitt's collection? The collection record for those pants takes a brief stab at explaining the significance of the acquisition but those 296 words only go so far. Every object that a museum has collected has something called a curatorial file. It is essentially the scrapbook that a curator keeps during the acquisition process. We almost never show them to anyone and when we do it's usually only one or two documents, a photograph for example, that serve as a kind of two-dimensional diorama for the object.

This is not a call for some kind of radical transparency in curatorial practice. That's an interesting conversation to have some day but that's not what I am proposing. Curators, like everyone else, should be afforded the space to work, to investigate, to noodle, in private. But there's a lot of room — a lot of stuff actually — between publishing someone's notes in the margins and distilling all the reasons why something was acquired in to short turgid micro-essays designed to fit on an index card taped to the wall.

Every single museum operating today has this problem. That's the bad news. The good news is that we all have curatorial files and in them is the opportunity to do better explain the meaning of objects in our collections and to explain the different reasons that different institutions acquire the same things. We can use those curatorial files to demonstrate the velocity of process, unique to curators and to institutions, that guides why we bother to keep any of this stuff at all.

For example, what does the curatorial file for the now-destroyed laptop that the Guardian newspaper used to store their copy of the Snowden files look like?

Here’s a thing: the visible and the invisible are products of the same belief system, and that is that all things are ultimately knowable. Wikileaks and the NSA believe the same thing: that if we can just bring all the secrets of the world to light, everything will be made good and right with the world. The museum and the software programme have the same essential ontology, that ordering things and structuring them in the right way will produce a representation so perfect we can build a whole culture atop it. But much is not knowable, and while the internet is perhaps here to reveal to us the vastness of what we do not and cannot know, the social and political philosophy with which it is mostly closely associated asserts not merely that everything is knowable, but that all things are knowable at once, and structures the world around this assumption, whether through surveillance, big data, the veneration of the market or the supremacy of the nation state. The curse of omniscience, once attributed to God, is now more tragically invested in the machine.

Update: Since publishing this blog post James Bridle's Hyper—Stacks project launched at the V&A and his write-up of the project is an interesting and worthwhile poking at some of these ideas. Go ahead. I'll wait...

A few weeks ago CBC Radio broadcast a documentary titled Objects of Burden. The documentary took its name from a project its subject, an artist named Raphaëlle de Groot, has been working on.

She invited people to donate objects that were full of meaning, that they couldn't just throw out, but no longer wanted. In short, objects of burden. In return, De Groot promised to take on each burden as her own. She promised to breathe new life into the objects, turn them into art.

Ignoring the ...turning things in to art bit this is kind of what museums do, right? Or maybe it's what we should be doing? Or maybe it's what people mistakenly assume we do? Maybe it's what they wish we'd do?

Okay, the Pen.

Raise your hand if you've ever been to a museum. Keep your hand raised if you've ever taken a picture of a wall label. Finally, keep your hand raised if you think that's worked out well for you. In all the time I've been asking these questions no one has ever thought it's worked out well.

Imagine if you never had to do that again. That's pretty much the entire reason the Pen exists.

Would that I could understand why YouTube's iframe code breaks the layout for my blog so here's a link and a screenshot instead...

I'd like to play a short video that we have on a continuous loop in the museum's lobby to introduce the Pen to visitors and help them understand what it is for and how it works.

The Pen is a large stylus (about the size of a long grease pencil) like you would use on a tablet or, in the museum's case, an interactive table. At the rear of the Pen is a small NFC antenna to read tags that have been embedded in all the object labels throughout the museum. Every label contains the unique ID for the object it describes. The Pen also has a small microcontroller and a storage device so each tag you read is saved to your Pen. You can view the objects you've collected on the interactive tables I mentioned or one our website at the end of your visit. Every visitor's ticket has a unique URL (or visit ID) that is paired with their Pen so once you transfer stuff off the Pen it's all available to you on the web.

If you want to keep taking pictures of wall labels that is fine but the point is that you don't have to in order to remember your visit, anymore.

The Pen was and remains a pretty massive undertaking. Museums don't really do their own product design, let alone try to tackle the realities of manufacturing. The Pen has been an idea that no one wanted to let go of for almost three years and a nearly all-consuming project for the last 12 to 18 months, depending on how you're counting.

We did not do this alone. We could not have done it alone. On bad days there is still the lingering question of whether we should have done it at all. I have never doubted that the project was worth doing even when it was not clear we could pull it off. In the end we re-opened the museum three months ago and only finally launched the Pen... well, yesterday. That delay has proven to be a blessing in disguise for a number of reasons but no one was happy about it at the time.

The Pen was not just a self-indulgent exercise in hardware manufacturing. We went to all of this trouble because we believe that it is the best way to achieve the goals the museum set itself when we closed three years ago. Personally, I would like to see an equivalent device in other museums because I think it's really useful but we built a tool for our museum and not for all museums.

So while maybe we shouldn't have had to take on the burden of creating the Pen from scratch what I find rewarding is the fact that we could even consider doing it. Five years ago none of this would have been possible.

A brief house-keeping note. The graphic above is missing one important name: Bloomberg Philanthropies. They paid for most of the Pen and aside from platitudes about thanking your funders this was a sufficiently crazy-ambitious project that they might have been forgiven for taking a pass. They didn't and that's worth calling out and saying Thanks to.

The Pen began life as a design concept, in other words a slide in a presentation. The implementation details were always tomorrow's problem which made sense for a time but with every passing tomorrow started to take on a greater urgency.

As forces in the Universe go tomorrow is pretty dense has a way of sinking to the bottom quickly and taking everything in its orbit along for the ride. The issue for the museum was how to short-circuit that inertia because it had become clear that the questions we were actively not answering about the Pen were going to blow back, sooner rather than later, on every other aspect of the museum's reopening.

So we taped an RFID tag to a whiteboard marker.

We bought some off-the-shelf parts that we knew were probably not what we would use in production and we taped things to them. We wrote some of the ugliest code you'll never see and used it to call our own API to update some of the ugliest webpages you'll also never see.

We did this so that the idea of visitors using a NFC-enabled pen in the galleries stopped being an idea and became something tangible. The problem with conceptual designs is that at a certain point they stop being devices for imagining possibilities and instead become a bucket for everyone's hopes and fears and anxieties. That tipping point is unique to every project but we had reached ours and the most important thing became to root the problem in a practical reality that we could use to make decisions about specifics rather than around them.

The earliest concepts for the Pen involved a passive label and an active pen, where all the computing power and other smarts where housed in the device. Our early prototypes flipped that model which also allowed us to play with the idea of replacing the pen with other objects like a big honking super-hero ring.

In the end we returned to the idea of passive labels and an active pen which led us to Sistelnetworks, in Spain who had produced a device called the vWand. The vWand is an inventory tracking tool that has no onboard storage and pairs itself with an Android device over Bluetooth and needs to be charged once a day so it's not really fit for use in a museum.

On the other hand it was pretty close to what we wanted, at least in broad strokes, they appeared to understand the death-by-a-thousand-cuts details of NFC and they had shipped molded plastic in a box with an FCC label so we got in touch with them. The black pen, to the right in this slide, is the vWand and the white pen is the first prototype of what became eventually became the Pen.

This is when the GE Design Council (which has a new name now but I can't remember what it is...) got involved. Full disclosure: Beth Comstock, the CMO at GE, also sits on the museum's board. She asked what GE could do to help out. We always understood that figuring out the electronics was only the first step. Or rather once it started becoming clear that what we were trying to do couldn't simply be bolted on, or in, to an existing product it became equally clear that we could use all the industrial and interaction design help we could get. When you consider that the Pen has no screen and that the only way to relay feedback to a visitor is through a small set of monochromatic (read: white, because power consumption) lights and a gentle vibration motor you start to get in to the weeds pretty quickly.

These are some of the early designs they did for us that never went very far for... a variety of reasons.

You can see the articulation of the museum's concerns in the evolution from the white 3D printed pencil on your left to the green marker on the right. There are three physical realities that have had an outsized influence on the final design of the Pen.

First, there is still a minimum width for NFC antennas. Anything smaller and they don't work or they don't work reliably. It's about 13mm. So, unless you're going to make a rectangular object that's the minimum diameter for any device before you have to increase it by however many millimeters (times two) the housing requires.

Second, the museum can't be needing to recharge the Pens every day. Or even every week. We have neither the time nor the staff. The pens need to be able to stay out on the floors in regular use for up to three weeks.

Third, the museum is housed in an historic mansion where space is at a premium and changes are not made lightly or at all. If you can design me a room that can accomodate a couple thousand micro-USB cables all plugged in at the same time then we can have a conversation about where it would go in Andrew Carnegie's house.

Which is why we chose to use AAA batteries to charge the Pen. Which is why we chose to use three AAA batteries instead of two because after we ran the numbers we saw that fewer batteries in each pen would mean we'd actually end up using something like an additional 50, 000 batteries every year.

GE did an incredible amount of work on the Pen. GE was also very upfront with us that they would not do the mechanical engineering, sometimes called design-for-manufacture or DFM, for their designs. It's not that GE couldn't do it but they are set up for producing wind turbines and submarines and we just don't fit into either category or scale.

A Matter Battle is the conflict between human intentions and the laws and behaviors of the physical universe. Material acts that are without intention, or where intention is purposefully exploratory such as drip painting, are not Matter Battles.

Matter battle is a phrase coined by Bryan Boyer in a 2011 essay. It turns out that making material things, things that have play nicely with other material things in the physical world, is still fantastically hard.

In 2015 we more or less have the software side of complex interactive systems under control. It's not without challenges but the barriers to implementation and participation have been steadily falling. We're not really there with electronics but you can feel that it's changing. It's hundreds or thousands of dollars to have your own PCB printed in small batches rather than tens of thousands. Making and producing an elegant and durable container for those first two systems remains so fraught with complexities and discouraging moments that it often seem beyond the means of most people, still.

This is especially true if you can't just burn the problems away with money which is what a lot of people do. The important thing to remember about the Pen is that it is not schwag. We are trying to build a sustainable piece of infrastructure that the museum can use and adapt for years so cost, over time, is a real concern for us.

It's getting better but there is a reason why so many ITP or other DIY projects are in Altoids tins.

It turns out the Pen is a pretty good problem-solving interview question. You start with two immutable facts of nature and a warning. Fact number one is that all capacitive styluses have a metal core or metal woven in to the sheathing. (Go back and look at the slide with the vWand cases — that's metallic paint on the tip.) That metal is required in order for the stylus to work. Fact number two is that metal is the enemy of radio frequencies (NFC). The stern warning is that if any point the person answering the question says I saw a thing... on 60 Minutes... about a guy in Shenzhen... then the interview is over.

Otherwise you just sit back and listen.

If they get far enough to figure out a design then you ask them how they'd power the thing. You can't really see it in any of the slides I've shown you but there's a button on the back of the Pen. That's the button which activates the NFC antenna because if it were always powered on the Pen would spend all day shouting HELLO? IS ANYONE OUT THERE?? in to the void and quickly exhaust its battery supply.

Putting the antenna on the back of a pen or pencil shaped object is a bit confusing for people so now you have the visual challenge of conveying to people that they need to collect objects by erasing them. Maybe in a world of ubiquitous sensor-phones and tablets this learned behaviour won't be necessary to challenge but we have a heterogeneous audience so it's a misunderstanding we have to deal with.

If you're lucky the interview candidate has read Timo's PhD thesis and knows that RFID and NFC have irregular fields, influenced by every spec of nearby metal, and that the antennae don't really work at perpendicular angles. A nice solid 45 degree angle is best but it turns out that holding a stick at a 45 degree angle and then pressing down really hurts your wrists.

Then you start tooling and co-ordinating all the moving pieces across an equal number of factories and their schedules and the combinatorial possibility space of the interactions between all those things means that there is still no way to figure out anything ahead of time.

So now the Pen is starting to push back against how the labels are designed and more importantly produced and installed. There's a little collect icon (a crosshair) on all the labels which is repeated on the Pen's button. We tell people to touch the crosshairs to the crosshairs but the reality is that the actual NFC tag is beneath the icon because that's where the NFC antenna in the Pen can see the tag when it's held at a 30 (not 45) degree angle.

Visitors don't need to know that but the museum does.

Did you notice that ink-y black tile on top of the label rail but beneath the NFC tag in my last slide? That's a thing we had to do because our label rails are made of solid metal. On top of everything else we've become dinner-party experts on a material called ferrite shielding used to distract all the magnetic interference caused by the rails themselves.

Physics...

We've had to build our own tool to write NFC tags. That's been a good exercise for us. The idea is that any staff member can look up all the objects on display by room number or by exhibition title and then update the tags one at a time. It's just a glue layer between a lot of built-in NFC functionality provided by the operating system (Android) and our own API. Besides giving us a real world use case to test our own API with it also forces the museum to keep its own data current and accurate.

And now suddenly the exhibitions team has to think about how they are going to get all those tags on to all those labels. The point is not that any of this is impossible but that none of it is obvious and that it starts to touch — to affect — every part of the museum's operations.

So, you've come to the museum and walked around the museum with your Pen collecting objects. Now what? How do you see all those objects? Until you transfer that data from the Pen to the collections website they are all trapped on your Pen. So we need to design and build transfer stations that allow you to tap your Pen and have unseen software and electronics take care of all the details.

Remember: Software and electronics are mostly solved (or manageable) problems. Case design, on the other hand... Hey look, we're designing and manufacturing another user-facing case!

We also need to pair Pens with tickets so we can distinguish one visit from another. These are devices meant for use by the visitor services staff and so they aren't anything the general public will see but that doesn't make them any easier. Actually, they've been an order of magnitude more complicated.

We've been working with Tellart on the electronics and the software for the registration and the transfer stations. As part of the original brief we all agreed that the ideal solution would be pair of simple and elegant objects that you could simply wave a pen over, or maybe touch, and all the details would be taken care of for you seamlessly behind the seamless seams.

In many ways we've gotten close to that with the transfer stations.

What we've come to realize is that this approach is not only unrealistic for the registration stations but actively counter-productive, in practice. Visitor services staff need feedback and something like an audit-trail that allows them to pair pens and answer three other questions and generally do sixteen different things at the same time.

So it's a user-interface and a error-reporting problem which really means it's a software-tooling problem for the museum: How quickly can we make changes and deploy them to production as front-of-house tells us what does and doesn't work? Every aspect of the Pen is brand new for everyone and there is no way to know in advance what will work so the most important thing is our ability to make changes.

The same is true of all the stuff running behind the scenes. Everything I've described so far was happening at the same time so there was not the luxury of waiting for one project to complete before starting another. All the projects were in process simultaneously, each pushing back against the other.

Before the re-opening, and before the Pen was announced, we used to tell people that when everything was done the building itself (the museum) would be the single largest consumer of our own API. The Pen is basically a fancy API application which has allowed us to build tools like the holodeck to test the basic functionality of the overall system independent of the final implementation details.

Did I mention that all the regulations ever written prohibit electronic equipment from being shipped with batteries pre-installed. It's always nice to imagine that these details are someone else's problem but they rarely are and so one day the entire museum staff got together and helped put batteries in all the Pens.

And I haven't even mentioned the interactive tables, yet.

I haven't mentioned that in addition to the Pens and the labels and all the related infrastructure the museum also sports a series of large interactive multi-touch, multi-user, tables designed to browse the things you've collected, related objects in the collection and to make your own creations as well.

I haven't mentioned that challenges of designing the tables themselves. The largest of the tables have 84" 4K screens (there are also 55" and 32" versions in the galleries) which literally didn't exist last summer.

I haven't mentioned the challenges of asking (forcing) multiple contractors to work together and to integrate both software and hardware components, all of which were in constant flux, across radically different schedules.

I haven't mentioned the challenges of asking (forcing) contractors to integrate with our own software; to understand that the museum's own API is the single source of truth and that yet-another black-box database to power an interactive application was not an option.

I haven't mentioned the challenges of digitizing (in some cases re-digitizing) the objects in our collection and requiring often extensive metadata clean-up in order to have something to display on the tables.

I haven't mentioned the challenges of asking an already busy curatorial department to also tag those objects, in non-expert terms, and to choose multiple related objects from the collection for every single object on display in the galleries.

I haven't mentioned the challenges of ticketing and related identity services. I haven't mentioned that one of the things the Pen forces is a collapsing of all the different databases, inside the museum, of members and donors and benefactors and so on. One consequence of the way the Pen works is that everyone is a visitor, is always a member, and over the course of their history with the museum might be a patron one year and not another. But their member-iness, that relationship with the museum, persists over time.

Maybe if we had all day I would talk about those things.

Instead I am going to suggest that you watch this video. It was produced by Local Projects who designed and built the software that runs on the tables. It's good work and the video offers a glimpse at the scale of everything I haven't been able to talk about today.

Here is another link to a video and article that was posted on The Verge yesterday. It's mostly more of me yammering on, like I am right now, but it's also full of good pictures of the tables and the other digital stuff that we built up around the Pen.

By now you may be asking yourself: Why would anyone choose to take on a project this big and this complex? That's a reasonable question. The museum asked itself that question many times during the last three years.

This is why.

Almost everything I've told you about so far today is stuff in the tip of this pyramid. I've touch on the API which is a little closer to the base of what we're doing. Mostly though I've talked about the things that we hope all of the foundational work make possible in the service of the museum's mission as a design museum.

It turns out that the Pen is in fact the minimum amount of infrastructure that you need if the goal is to enable some kind of meaningful recall for a museum visit. The point is not to provide visitors with a Pen experience but to offer them a tool that is quiet and polite and allows them to, literally, touch the objects as a way to remember them. To provide them with something less-shit than taking photos of wall labels. To provide them with a way to come to the museum and have a heads up visit confident that there is a way back in after they've left the building.

There is one other thing that, in time, may prove be the most important change to come out of the last three years: We updated the loan agreements.

We changed the loan agreements to state that the museum reserves the right to display the fact that an object spent time with us and to display the images of those objects on our website and in our galleries. Forever. If you're not a museum person you may be staring at your screen right now wondering what the fuck I am talking about. Like, specifically, why this is a big deal. That is the correct response.

Pretty much every other loan agreement ever drafted between two museums or a museum and a private individual states that lender retains all image rights to the object being lent. Which is fine, in principle. In practice though it's created an environment where even if a museum enjoys a limited period of use the uncertainty around the licensing of that imagery after the fact means that it's easier to throw up our hands and despair the situation than to look for a viable alternative.

The problem is this: We tell visitors that it is important enough for them to travel to our musuem to see something in person rather than simply looking for it on Google. We tell them it is worth their time and expense and then we pretend as though it never happened.

Which is insane. It's flat out insane. Not to mention wrong. Also stupid.

So we've stopped doing it. We're not going to start making mugs and ties with other people's collections but we are going to assert that their thing was in our building for a while.

This is why.

This is also the first (and maybe last) time I've ever used fancy-pants transitions in a presentation. What's supposed to happen here is a slow and graceful panning from the top to the bottom of this screenshot because it is 5, 672 pixels tall. You can see the original over here (7MB) if you're curious.

Note: We do not actually have Mike Migurski's 2007 prediction about the collapsing distinction between math and design in our collection. We should, though.

Every act of collecting an object or of creating something on one of the interactive tables has a permalink, a stable and permanent URL that is part of the collections website.

By default those pages are private but you can make them public if you want and share them over Twitter or email or whatever you want. It is not for the museum to decide why you should collect something or what you might want to do with it afterwards. We need to have the confidence that people will figure out what to do with this stuff, on their own terms.

The challenge right now will be in conveying to people that a stable and permanent record of their visit is possible and to prove to them that we are serious in our endeavour. Talk is cheap, I know, but we are the Smithsonian and we aren't going anywhere any time soon and so we have the privilege and luxury of doing this and doing it for real.

I think the most important, most valuable, moment will come not during a visit to the museum or immediately afterwards. It will come weeks or months later when a person is out at the bar talking to their friends and will remember something they saw during a past visit. You know that moment where you're looking at your friend saying this like The thing... it was... yellow...

Let's be honest: You are straight-up fucked if you then try to search for that thing on a museum website and doubly-fucked if you're trying to do it on your phone. We should all strive to make that experience not suck but for the time being it does. If instead a person can remember that Oh yeah, I was there in October... and there's a way to find the object quickly and easily then two things happen:

1. They can actually find the thing they're talking about and not have it be a proxy object for another of life's annoyances.
2. They can put their phone away.

Imagine if you could take a museum for granted that way. Not in a creepy or selfish way but in a way that allowed you to think about it as a resource, with the patience to always be present. Imagine what it would mean for a museum to have the infinite space of everything to the right of a permalink's URL at its disposal.

It's not a permalink of the object (they already have their own permalinks) but a permalink of your having collected that object during that visit and these are the places where visitors and the museum together might actually explore what it means to better share an understanding of an object beyond a 75-word wall label. There is a fantastic amount of learning and writing that has produced about the objects in our collections over the years but almost no one, outside the hula-hoop of professional disciplines, ever sees it.

These, we hope, are the places where we might start to change that. These are the places where someone might finally read the 10,000 word essay about an exhibition in the comfort of their living room or even just on the subway ride home after their visit. These are the places where we might start to find a way to make the curatorial files I mentioned earlier an active participant in the collection.

There is a saying in some programming circles that If you launch a feature or a product without bugs then you've waited too long to launch. If this were true of all software projects the world would be a worse place but it's not a bad rule of thumb for many things.

You may have noticed a few oddities as the screenshot of the visit page scrolled by, namely the pixelated hand and the Google Street View cat. What this means is that our own collections database (the canonical source of truth) is out of sync with reality. Two of the objects on display in the galleries haven't had their images made public and one of them hasn't been made public at all.

The reality of exhibitions though is that they a work-in-progress almost to the exact moment that the doors are opened to the public. As a consequence museum practice has evolved in such a way that our collections databases are often updated only after an exhibition has launched. We can't do that anymore.

Like all the other things we will figure out the details over time and while we do it offers another opportunity to think about how and what we show visitors when a robot thinks they should only see silence.

We also since replaced the Street View cat with Bao Bao.

Yes, I know. It's a picture of a wall label. It was taken at the V&A's Disobedient Objects exhibition where, no joke, I was not allowed to take photographs...

In the end I think the hardest part of this project for the museum will be being patient and in measuring success over the long-term. Some people will see and immediate and personal value in what we're trying to do but it would be unfair, and unrealistic, to demand the same of everyone else. People have busy, complicated lives and it sometimes takes people a while to warm up to an idea. Our disposition, our super-power, as cultural heritage institutions is that we have time on our side.

We should learn to share it with those who don't.

Thank you.