The hard work of imagining, ThingsCon 2020

16.00, Friday 11 Dec 2020

I presented this essay as part of ThingsCon 2020 on 11 December, 2020. The week-long virtual festival was also the launch of this year’s Responsible Internet of Things publication, so I decided to speak about imagining futures - dystopias and utopias. As a talk, the essay was shortened and I also used slides. What follows is the long-form version.

Dentistry, 3D printing, and the Gartner Hype Cycle

I want to talk about the Internet of Things and how we build the future. The theme being, of course, as this is the event, responsible IoT.

The Internet of Things first appeared on the Gartner Hype Cycle in 2011. (Source; see linked spreadsheet for data.)

Also that year:

  • Image Recognition. Well, that’s just how computers see now.
  • Also QR codes. I was rooting for you, QR codes. It only took a global pandemic to have me scanning codes every time I leave the house.

It’s interesting the convoluted route technologies take to adoption, and the effect they have, and to think about the Internet of Things, IoT, in that context.

3D printing first appeared in 2007. It had a 4 years head start.

We got our first milling machine in 2006. This was at Schulze & Webb, the studio that later became my old company BERG. It milled blocks of chemical wood, and produced a kind of dust that gave us “tight lung” as we called it then because we were young, but will probably take a year off my life when I’m old. It ran overnight to make any kind of shape. Resolution, and I’m guessing here, probably around half a millimetre.

Now I went to the dentist the other day to get a crown replaced. My mouth was scanned and the tooth designed with a handheld photogrammetry device and touchscreen 3D software on a terminal right by the chair.

It took 7 minutes to mill my new tooth on the machine in the basement. The milling machine has an accuracy measured in microns: about a thousand times more accurate than our old miller.

Even last year, making and fitting a new crown was a 2 week job requiring a specialist lab. Now my neighbourhood dentist can do it, and I was in and out in less than 90 minutes.

So the way 3D printing has come into the world is what I would call: same but faster.

Faster teeth. Faster prototyping. Faster tools for injection moulding in factories.

But we haven’t seen the society-wide impact that I think some of us were expecting when 3D printing first appeared on the Hype Cycle. Supply chains haven’t turned into supply webs. Manufacturing hasn’t become local, with mini factories in every neighbourhood. We don’t print our phones or print our shoes or print our breakfast. We still have mass production and mass consumption and mass marketing.

Don’t get me wrong. I hate going to the dentist. 3D printing means I can go to the dentist for less time. I am delighted. But what we got is nowhere near what we imagined.

So is the Internet of Things more like what we got with 3D printing, or more like what we imagined we’d get? Is it same-but-more-efficient, or is it transformative?

The Great Inversion is midway done

I think about the Internet of Things as the Great Inversion.

It used to be, before the Internet of Things, say before it appeared on the Gartner Hype Cycle, before 2011, that the computer was contained in the world. There was the world and it contained people, and forests, and cities, and shoes, and feelings, and all the rest. And one of the things contained by the world was the internet.

Because of the Internet of Things, this situation has inverted, it is inside out. There is now the computer, and one of the things that it contains is the world.

Robotics are the hands and feet of that inversion. Computer vision – that’s the eyes. But IoT, it’s the connective tissue. The wiring.

I would say that the Great Inversion is currently midway done.

There’s a short story written by Paul Ford way back in 2002 called Robot Exclusion Protocol. It is very short, 254 words. It’s about Google. Here’s how it starts.

I took off my clothes and stepped into the shower to find another one sitting near the drain. It was about 2 feet tall and made of metal, with bright camera-lens eyes and a few dozen gripping arms. …

“Hi! I’m from Google. I’m a Googlebot! … I’m indexing your apartment.”

That’s the era we’re in at the moment. Indexing. Ingesting. Eating.

With IoT we’ve got industrial IoT with sensors in factories, and we’ve got connected cars, and we’ve got voice-controlled gadgets.

And what it means it that the physical world is now subject to all the winds and forces of the internet. Those search engine index, those trading algorithms; it’s subject to analytics and automatic optimisation and machine learning, and all the rest.

I don’t see it slowing down. The economic imperatives are too strong.

Whatever happens to musicians happens to everybody

What happens when software eats the world? Bruce Sterling has a line about this. He says, Whatever Happens to Musicians Happens to Everybody.

He talks about a collapse in genre diversity, and the distributors taking control of the economics, and go-it-alone creators. And his point is that you can see the same thing happening in newspapers and fashion and whatever. It happened first to musicians.

Because of the Great Inversion, the world is part of the internet.

So I’d like to generalise Sterling’s Law to this: whatever happens on the internet will happen to the world.

And there are many good things about the internet.

But the internet in 2020, well, we’re not in a good place. Despite the idealism of those who wrote the RFCs when the internet was in its infancy, and the good intentions those who who were at the vanguard of Web 2.0 when the internet came into the mainstream.

Not in a good place at all.

Platform capitalism and consumer farming

Like any complex system, the internet comes with internal forces that shape its evolution. Forces, tendencies, gravities. Call this internet realism. I call these forces “logics” because they are directions that just make sense within the context of the internet.

One is the logic of platform capitalism.

Platform capitalism is a term invented by the economist Nick Srnicek. His book is great. He uses it to label the operating model of many Big Tech corporations, and he points out that they work like this:

  1. Operate a marketplace that brings buyers and sellers together. For example, book sellers and book customers. Or drivers and passengers. Or advertisers and browsers
  2. Collect data - any data - all the data - that can be used to drive marketplace activity, either by pushing transaction volume or marketplace size.
  3. As the marketplace grows, it wins out over other, alternative marketplaces, reducing competition. And the data capture basin increases too, enabling further growth. As part of this step of the operating model, data capture always increases.
  4. Repeat and grow.

You can see it in action with, say, Facebook. All my friends are there! It’s free! There’s no excuse for me not to be a participant in this marketplace. Yet the data which is gathered is used to drive my marketplace activity – my clicking on ads and my purchases. From the advertisers point of view, there’s no ability for them to opt out either. And the ads are priced at just the level where they would be foolish not to participate – but where Facebook can keep as much margin as possible. The data they gather lets them know exactly what this level is.

This is a hungry logic. It’s expansionist, and there’s no room to realistically consider alternatives. It just makes sense.

What happens when the logic of platform capitalism meets the Internet of Things?

We see glimpses of that with Uber, and their carefully priced marketplaces which capture drivers into vehicle rental and subsistence income. And we see glimpses of that with the Amazon Ring doorbell and data gathering on the street.

But let’s take it an extreme. I can imagine a free apartment where everyday activity is monetised. A free house that comes with a bundled app store of Amazon Dash-style subscription purchases for cleaning products, and food, and clothes, and rental furniture; all carefully and automatically optimised by monitoring usage through connected cameras and sensors.

This is a dystopia where humans, you and me, are farmed as consumers, by platform capitalism. We never own, we pay rent.

It could be built today. It’s just that nobody’s gotten round to it yet.

So that’s one example of where that particular logic could end up.

To the computer, it’s just another device

A second logic of the internet is that of abstraction.

Ted Nelson invented hypertext and was one of the first to really probe what it meant to use computers for creativity. He’s a visionary. In his book Computer Lib (1974), he said this:

Whatever it may do in the real world, to the computer program it’s just another device.

And this is the amazing thing about computers and the internet – the computer sends data, and it could be showing a few pixels on a screen, or it could be driving a probe on another planet.

Or it could be transferring Bitcoin. A Bitcoin transaction takes as much energy - and therefore as much carbon - as used by an average British household in two months.

These all have the same weight to the computer. And so they all weigh the same to us.

The logic of abstraction is neither good nor bad, it’s just the way computing works. Everything gets abstracted. Computer scientists have a name for when you can tell what happens the other side: they call it a “leaky abstraction.” It’s something to be avoided.

But in the real world, when consequences are hidden, situations tend to be abused.

And what happens when the real world is, well, just another device, thanks to IoT?

Well, you tap an app on your phone, and you call a car. The driver rents their car, is paid below minimum wage, has no savings. They’re an independent contractor so they don’t have employment benefits. At some point the workers will be automated away. The company is structured so that they are, in the parlance, “asset light” - but also so that they can’t be held accountable.

We, the tapper of the app, are insulated, because of the logic of abstraction.

It’s what Peter Reinhardt first referred to as Below the API versus Above the API jobs.

Living above the API, we order groceries, interact with customer service, live our lives one step abstracted from the people with whom we share a society. And below the API, wages are squeezed, people made to compete with robots, and inequality grows.

It’s hard to argue with the logic of efficiency and automation. That’s what makes it a logic. The logic of abstraction is what makes it hard for us to even see it happening.

The logics of monocultures and overextended complexity

So there are other logics of the internet. The logic of monocultures that create attack surfaces for state-sponsored cyberwar. The logic of overextended complexity that grows till breaking point. We’ve seen high-frequency trading algorithms cause flash crashes in the stock market, from just that logic of complexity – what happens when we have a flash crash on a highway of autonomous cars filled with school kids and commuters?

But enough dystopias.

Dystopia is an extrapolation. Utopia requires discontinuity

The future I want embodies different values.

I want my home to be voice controlled – but I want it without centralised data capture.

I want car-sharing schemes and last mile delivery – but I want it to operate through mutualism and co-operatives.

I want home security drones – but I want them to make their domestic visual index available to a private app so I can text-search my bookshelves for that book I can’t lay my hands on, rather than them using it to train their ad targeting A.I.

Privacy. Agency. Mutualism. Equality.

I’m not saying these futures are impossible, but they feel hard to reach from where we are right now.

Because honestly, it feels like I don’t get to choose, we don’t get to choose, those futures.

The system that does get to choose is the system that surrounds our tribe of designers, technologists, and founders. It’s all those adjacent tribes that never get to see the idealism and the intentions. The marketers, retailers, supply chain experts, risk assessors, the MBAs, policy-makers, and so on. They don’t get to see the vision; they have to follow the well-trodden path. They follow the logics. The logics are the same old business models, the same old ways of capturing attention, the same old methods to build platforms.

Dystopia is the extrapolation of the same old, same old.

But utopia is a non-extrapolation, it requires a discontinuity. It requires all these different tribes to choose to do something different, at great risk to their careers and livelihoods.

If they’re going to do that, they all need to be shown that something different first, and shown how it’ll work.

And you know what, I think that’s our job.

We need to do the hard work of imagining

Sir Terence Conran: The designer’s job is to imagine the world not how it is, but how it should be.

I’ve lost the habit of imagining utopias. Perhaps we all have.

Today, right now, we’re in what architect Bryan Boyer calls a “vision vacuum.” I don’t know why, but imagination about positive futures is scarce right now. And in the absence of a compelling narrative, the same old, same old wins by default.

Narratives? Fiction?

But we don’t need just design fictions. We need business model fictions, engineering feasibility study fictions, interop protocol specification fictions, investment return fictions.

I’ll give you an example. I’m a proud member of the British Interplanetary Society which is, on one hand, a talking shop. But on the other, has been a significant part of the conversation, for 87 years, through engineering feasibility reports. The society’s engineering design for a probe that would go to another star, a study called Project Daedalus which ran from 1973 to 1978, has helped make the idea of interstellar travel more believable. Concept ships in research and fiction are based on the Daedalus designs; the Daedalus report helps identify what the actual problems are and where R&D should be targeted.

So I guess what I’m asking for is a different kind of think tank, not one that works with recommendations and reports and regulation, but a new think tank that trades in politically opinionated, worked examples that demonstrate, demystify, and de-risk.

The objects and systems must be plain, easy to understand, and embody our values. If you asked me now where to start, I would start with worked examples for:

  • Zero-user-data connected products (related: I wrote about voice control for everything, but could we sketch a business model?)
  • Open file formats for service portability (related: I wrote about interop for video calls, but could we design the protocol and code up a reference app?)
  • Software-enabled cooperative corporations (related: I wrote about software-enabled co-ops but could we ship a co-op as a Shopify plugin?)

We need boundary objects that transcend language and can translate across the different tribes.

And the marketers, retailers, supply chain experts, risk assessors, the MBAs, policy-makers, and so on, if we can indeed make tiny, proof of concept, real versions of these futures, bonsai tree utopias, made out of spreadsheets and simulations, if we can speak to them in their words, they will nurture and help those futures grow, I think there’s space for that.

I think I need to - we need to - imagine utopias again, and we need to demand and create demand for them, and we need to articulate them in great detail.

But we need to be the ones doing the hard work of imagining a responsible future for the Internet of Things. Because it’s no-one else’s job to do so.

Follow-up posts:

If you enjoyed this post, please consider sharing it by email or on social media. Here’s the link. Thanks, —Matt.