Categories
Net Culture Personal Publishing Social Software Technology

How Threads will integrate with the Fediverse

This is an exceptionally long post detailing pretty much everything I learned at an event shortly before Christmas at Meta’s offices in San Francisco. I’ve been delayed in writing it up because of traveling back to the UK for Christmas and other commitments – and because I wanted to capture everything. It’s roughly written, and I’ll probably edit it a bit after posting. If you have questions or comments or want me to clarify anything, DM me on Twitter @tomcoates or email me at tom [at] plasticbag [dot] org.

Just before Christmas I was lucky enough to be invited to a Data Dialogue event at Meta’s offices in San Francisco. The event was designed to reach out to people in the ‘Fediverse’ community, tell us their plans for their product “Threads” and get a bit of feedback about the policy and privacy implications. 

Since that meeting, Mark Zuckerberg has announced the first part of the Threads roadmap – making it possible for people to see Threads posts within the wider Fediverse. Given that, I thought maybe it would be a good time to write about the other things I learned and some of the feedback we gave the company.

What is the ‘Fediverse’

For those of you who haven’t been keeping up, the Fediverse is one approach to the question, “How can we have one (or more) social network(s) that no one company owns, for which anyone can make a client or a server, with all of them interoperating as seamlessly as possible so that they’re understandable to people who aren’t terminally online”. 

The main Fediverse approach is through projects like Mastodon – which are effectively small, local social networks that can be hosted by an individual or company, but whose users can still communicate with — and reference posts and people — using other similar networks. Most of these products are built with at least some reference to the ActivityPub protocol co-written by Evan Prodromou.

There are other approaches to this idea of a ‘public’ (that is non-privately owned) social network system or protocol. Some run on crypto tech, where people run ‘relays’ (some of which generate crypto currency for the people who run the servers in the process) but the individual user completely owns and maintains their own identity and isn’t ‘hosted’ as such. 

I co-founded a company a few years ago –  funded by Bloomberg and other VCs – focused on one of those (built on the secure scuttlebutt SSB standard). We made an iOS client called Planetary. Since I left the company it has changed its name and pivoted to another protocol. That’s why I was invited to this event. I’m not going to talk about that much, but feel free to ping me if you have any questions. 

Anyway, there are lots of reasons why people should be switching to the Fediverse – among them:

(a) that one company should not generally be the main arbiter of what is acceptable speech for half the population of the planet;

(b) the general public should have the option to communicate with their friends (or find out information) without having that experience meditated by or optimized’ by algorithms generated by other people;

(c) there are reasonable questions that can be asked about whether or not a space entirely owned by advertising-focused companies can build products that aren’t socially corrosive or promote conflict and polarization.

However, there are also lots of reasons why people tend not to switch to the Fediverse – it can be challenging to understand so the process of using it presents a little more friction to the general public, some of the clients can be a bit clunky, and it’s often unclear how individual products within the space can support themselves financially. That’s why despite several big spikes in people leaving Post-Elon Twitter (I will not be calling it ‘X’) to join (in particular) Mastodon, the number of active users for the Fediverse has generally stayed in the low millions of people. That’s about 1/2000th the volume of the people who use Facebook/Instagram/WhatsApp etc. in any given month.

‘Threads’ and the Fediverse

So it was both very interesting and also a little alarming when Meta announced Threads around eight months ago – and at their launch they made it clear that their goal was for it to be part of the Fediverse. 

It was interesting because it seemed to indicate that something was finally changing in this space and that we could look forward to a world in which social networks operated a little bit more like e-mail – ie. used by billions of people, not owned by any one company, where you could choose your provider, but still connect with the entire world of other people.  

And it was a little alarming, because the current Fediverse is mostly enthusiasts and utopian individuals operating in a mostly non-corporate environment, with few (if any) algorithms and little (to no) advertising. It’s currently a space where people don’t generally have to worry about the billion-people-impacting, market-driven and perhaps dehumanizing decisions of massive companies or the fetid whims of asshole billionaires. That has tended to make the spaces much less corrosive, far less aggressive and really quite pleasant to be in.

It’s not unreasonable to wonder if such an environment can withstand the arrival of a social giant.

Anyway, after some initial excitement and dread, after eight months the Fediverse community had started to calm down a bit – mainly because it seemed like this integration was never going to happen. After all, Threads has had a very successful launch – with around a hundred and sixty million people signing up over the last few months. It is a highly active space, and its active user base is now a hundred times the size of the Fediverse with which they had claimed to want to connect. Had something changed? Was it all a bit of corporate flim flam? Was it just an attempt to market it as a more palatable and distinct alternative to Twitter that they really had no intention of following through on? Or had they maybe just changed their minds?

Well, I can report that the answer is no. They have not changed their minds. They seem to be very keen to continue to integrate Threads with the Fediverse. And at least superficially they seem to be attempting to do so carefully and in good faith. 

I have some more behind the scenes stories further down this post which I’ve heard from various players on the edges which might explain some of the motivations at the company, plus a bunch of speculation from other Fediverse attendees. And I have some concerns and questions about what they’re doing and how – both in terms of the impact I think it could have and also in terms of how it will be received by the community more widely. I also have a number of significant concerns with the Threads project itself. 

But I can report that in my opinion the teams building it and the integration seem to be decent people, trying to build something they’re excited by, wanting to be part of something new and truly federated, and wanting to be respectful and careful about how they do it. And whether or not you think their arrival in the space is a good thing, that apparent good faith and care has mitigated at least some of my concerns. 

Okay, so let’s get started with what they announced. One (hopefully final caveat), the whole thing was run under Chatham House rules, which means that I can talk about everything that happened at the event but I can’t ascribe what was said to specific people without their explicit permission. To anyone else reading this who was in attendance, you should feel free to quote me on anything I said. If you’re comfortable being attributed for anything below, then let me know (via Twitter or e-mail tom [at] plasticbag [dot] org and I’ll amend the post accordingly. 

The Product and the Roadmap

The first part of the session was focused on what their goal for Threads was and what the roadmap looked like. They started by stating the product was “a text based app for public conversation and to share your point of view on real time events” (so effectively a Twitter-clone, which we knew), where you could have “productive conversation and tune out the noise” and that it was important to them that it was “open and decentralized”. They seemed quite committed to the latter, explicitly saying things like “Threads will help people find community, no matter what app they use” and “If you don’t like the rules we are enforcing on our server, you will be able to take your followers elsewhere”.

I have some thoughts on that, which I’ll talk about later in this piece, but before that let’s talk about the roadmap they laid out, which is as follows:

December 2023 – A user will be able to opt in via the Threads app to have their posts *visible* to Mastodon clients. People would be able to reply and like those posts using their Mastodon clients, but those replies and likes would not be visible within the Threads application. Threads users would not be able to follow or see posts published across Mastodon servers, or reply to them or like then.

• Early 2024 (Part One) –  the Like counts on the Threads app would combine likes from Mastodon and Threads users

• Early 2024 (Part Two) – replies posted on Mastodon servers would be visible in the Threads application

Late 2024 – A “mixed” Fediverse and Threads experience where you will be able to follow Mastodon users within Threads, and reply to them and like them

• TBD – Full blended interoperability between Threads and Mastodon

Now there are a few bits in here I think are really interesting. Many of them make a lot of sense, but I still think will be controversial.

The first is how long this integration is going to take and where they’re starting work. I imagine there will be a bunch of people out there who think the early stages above feel like Facebook is ‘engaging’ with the fediverse in a pretty selfish way – ie. the first stage seems pretty one-sided, with Facebook pushing their content into other people’s servers, but not reciprocating.

This feels neither completely fair nor completely unfair. My first sentiment was similar – to what extent is this an integration rather than a colonization? You could view this as Facebook attempting to erase the fediverse – to take it over. After all, it’s unclear what the fediverse gets from having hundred million Facebook users pushing their content into their space without any ability to seriously reply and engage with them.

I still think that this first stage is likely to be the least popular and will drive the most discussion of Meta and whether they’re engaging in hardcore extractive and exploitative capitalism. I suspect many in the Fediverse will find this first move pretty repulsive.

On the other hand, I’ve worked in large companies. And I’ve worked with privacy and policy teams in complicated new areas (I built a product called Fire Eagle a million years ago which was pretty much the first one to handle user location data, and that triggered a good amount of terror among Yahoo’s privacy and compliance teams). 

And I’ve also had a (fairly sporadic and not always effective) place on the advisory board of the UK’s Open Rights Group. I know how hard this stuff can be and the dangers in the data and the space. Meta will have to deal with a lot of regulatory questions and privacy concerns before they can launch anything, they will have to figure out how they can engage with content and users who never signed up to their TOS and privacy policies, they’ll have to figure out what impact that will have on their other systems and business units, and they’ll have to develop a sense of the issues that campaigning groups, the FCC and the EU are likely to have with this new development. All of this stuff is a Big Deal. 

For me it makes sense they’d take this stuff slowly and carefully, and it also makes sense that they’d start off focused on what functionality they can offer their users who have signed up and opted in, before they start confronting the larger issues. It’s really the only way they could progress. And I guess they’ll just choose to suck up any negative response they get as a result.

I want to be very clear here – whether or not the way Facebook/Meta/Threads choose to handle this ends up being ethical or appropriate is really anyone’s guess at this moment – and I’m sure a bunch of people reading this have their suspicions that it won’t be. However—ethical or not—I believe they’re going to be very focused on being legally compliant and very conscious about avoiding (as they see it) the threat of further regulation. People often think large companies are more cavalier with the law than small ones. I can assure you in my experience exactly the opposite is the case. Large companies are mostly much more cautious about breaking the law, and instead invest much more of their money in trying to change the law in their direction. But that’s a story for another time.

Back to Meta’s roadmap: this project is also intensely technically difficult of course. It’s worth remembering that they didn’t build Threads out of pre-existing Mastodon open source code and they didn’t start with ActivityPub as a basis. They built it out of the Instagram codebase and community with a view to expanding that one network into this new parallel open and distributed space. As such, a bunch of core concepts and technical decisions are not directly and immediately compatible and will have to be rebuilt or redesigned to connect amicably. 

By the way, as someone who has built large products for a number of companies—including the BBC, Jawbone, Nokia and Yahoo, and run two start-ups—I have to tell you based on my limited knowledge at this point I think this roadmap is probably wildly optimistic. But I guess we’ll see.

Who was in attendance

Before I continue, I want to give you a sense of the people who were at this meeting. If I had to guess I’d say there were roughly twenty people present, falling roughly into three chunks – the first third were representatives of the Threads team, the second a group of legal, privacy and policy representatives from Meta and the rest of us were sort of roughly ‘representatives of the Fediverse’. 

As is probably obvious from a community that is specifically and self-consciously uncomfortable with monolithic organizations and wants to find a way for lots of smaller groups to cooperate, the Fediverse group was not particularly unified in our responses, nor were we representative of the ecology as a whole. There are a thousand projects and hundreds of talented and interesting people with more or less impact on the space. 

Still, some of the people present did have very, very really deep and long-standing engagements in Fediverse and ActivityPub projects. One of them said that of probably the ten most significant people trying to corral the space at the moment, three were in the room. I want to make clear that despite my deep interest in and work in the area I’m almost certainly not one of those three.

So, both a pretty serious group of people but also not representing the full range of opinions and views in the space. As ever, none of us can speak for the whole community. In fact at this point that would slightly miss the point.

Why are Meta doing this?

Anyway, despite our different perspectives, one question was clearly on everyone’s minds – Meta had talked about what Threads was and made it clear that openness and interoperability were key to the project – but they hadn’t talked about why they were doing it?

What on earth was motivating them to make this thing so – in theory – open? 

Their answer was that they simply felt it was the direction of travel for ‘social’ generally – that the area had been growing steadily, particularly post-Elon’s takeover of Twitter, but that they’d also had a lot of conversations with high profile people who build communities on their platforms and they were increasingly uncomfortable with Meta or Facebook or Instagram effectively owning their followers. They were looking for the ability to know that if they needed to they could move elsewhere. 

I’ll be blunt – I didn’t find this enormously convincing but it was interesting and I’m sure there’s some truth to it. It just didn’t feel like the whole story. We asked about the business side of things and they said obviously they were a business and moreover an advertising business, so probably that would be the way they made money in it long-term, but it wasn’t happening for a while. The representative said something along the lines of, “obviously we have a business model, and it’s fair to ask how this squares with that; the answer is we don’t actually know; there will likely be ads but not in the near future.”

After the event many of the Fediverse representatives speculated about other motives. They varied between the following:

1) Meta thinks Twitter’s part of the zeitgeist is important and powerful and are interested in that space – and they’re following Google’s response to the iPhone by promoting an open competitor they can benefit from;

2) Meta is concerned about greater regulation and are building out a space that perhaps they can still dominate but which they can make absolutely clear remains open, in order to shut down arguments (particularly from the American right) about how they’re censoring conservatives (they can move elsewhere) or antitrust laws (we’re directly creating an open environment where people can switch easily);

3) Someone down in the hierarchy doing a PM job just added in Fediverse support as a line item in a pitch deck to act as a differentiator and it’s just risen up through the ranks somehow surviving each time because many people simply didn’t know what it was. And now they have to build it; 

4) Mark Zuckerberg just hates Elon and is just doing everything he can to destroy Twitter.

I have absolutely no idea which one—or combination—of these is most accurate, but I can report one interesting story that I’ve now heard from two separate sources, one in attendance at the event and another friend in another part of the industry. According to both of them, at a hack day inside Meta someone presented this concept and a rough working prototype and said it was really interesting and exciting and that they should work on it and that it should totally be open, and that person, bizarrely, was Mark Zuckerberg himself. I have no idea if that’s true, but it’s certainly interesting and it would explain a lot of the enthusiasm for the venture inside the organization.

Anyway, I don’t think that’s necessarily incompatible with the motivations above, but I did think it was interesting. Come to your own conclusions, I guess.

(A note on that: I have many issues with Mark Zuckerberg’s approach to things—he’s definitely focused on business as the first priority and social responsibility is … well, it is somewhere on the list I’m sure—but he’s an intelligent man, and an engineer, an investor in some early decentralized and distributed social tech and someone who attended, with me and many other peers, social computing FOO Camps, organized by O’Reilly Media in the early days of the social web. He knows about ActivityPub and all of these open standards and has done for years. So if he’s the one who was pushing for it – I guess I’m not surprised! What’s more interesting – and actually quite excitingis that he thinks this might be the future of the social web!)

Godzilla & the Fediverse

Now, I mentioned above that the people we met at Meta seemed like decent, well-intentioned people attempting to do the right thing. However, this may not be enough to be a ‘good citizen’. And to understand why I think it’s worth talking briefly about the scale of the various parties. 

The community that Threads is planning to participate in is that of Mastodon servers federating with one another via Activity Pub. The estimates of this community are that there are about 9,500 separate mastodon instances participating in this ecology, with roughly 1.5 million Monthly Active Users (MAUs). This is a fairly substantial number but of course it pales in comparison to Meta more generally, which has closer to three billion active users. Or to put it another way, Mastodon users represent about 1/2000th of the number of people using Facebook/Instagram/Threads/WhatsApp etc. worldwide.

Threads itself has only been around for a few months now and it still towers over the rest of the Mastodon community in terms of users. It’s based on the Instagram user base, and Instagram users can opt in to use Threads with a single tap. Because of that—as of a recent earnings report—Meta can currently claim around 160 million total users and about 100 million MAUs for Threads alone. So, again, maybe we shouldn’t be thinking about Threads ‘integrating’ with the fediverse and instead think about Threads attempting to engage with the Fediverse without entirely crushing it in the process.

Effectively you can think of the existing Mastodon / Fediverse community as a pretty decently sized US city of people, with each server being a separate building within it. Meanwhile, Threads is an apparently friendly version of Godzilla, hundreds of times as tall as the nearest building, wanting to say hi to people, but every time it makes a move, there’s a risk its tail will kill thousands.

Or perhaps it’s more like the spaceships from Independence Day, only—you know—trying to be nice. It’s not hard to imagine that however well intentioned they’re going to be, their presence is going to be absolutely enormous and potentially catastrophic. 

Anyway, I mention all of this scale because I think it’s really important to keep it in mind when thinking of other issues that came up at the event – including how content moderation is likely to work, what kind of public education work they need to do, and what the rest of us need to think about in order to make sure that their arrival actually does open up social media, rather than completely destroy the independent communities that already exist. 

Content moderation

So let’s start off by talking about Content moderation. This was an issue that came up regularly during the day. Clearly, Meta moderates its own content on its own servers and will reserve the ability to remove that content and ban any user it wants. That is unlikely to change. 

However, the Fediverse presents some interesting challenges here. The goal of integrating with the fediverse is specifically to have Meta users’ content appear in someone else’s mastodon instance, and vice versa

This definitely appears to be an area that is causing them some concern and confusion – and they don’t seem to be 100% clear on how to handle it. For a start, all their users opt in to using Meta, but third party users do not. Nonetheless from the discussion it seems obvious that they’re going to have to reserve the right to exclude content and users from being cached on their servers or from being visible in their app if it breaks their rules even if it originates in another Fediverse instance. The same will presumably be true the other way around – individuals in public will be able to ban Meta’s instances from engaging with their instance or ban or block abusive Meta users.

One conversation that emerged around this was whether there was a way that Facebook could usefully open up these decisions more widely to benefit the larger community. This is an idea I’ve been keen on for a while – that an entity that is doing moderation work could open that up as a service which third parties could subscribe to. This might be a way that smaller instances could actually benefit from Meta’s presence, resulting in a better moderated space overall.

Of course the negative side of that is that Meta is not actually known for being particularly rigorous in their moderation, and in fact only has around one moderator for every 100,000 users last time I checked. Plus of course, this would entrench their already vast power as the arbiter of what is acceptable or unacceptable speech on the internet even further than it already has been. So, there are some significant risks.

Bluesky’s content tagging was mentioned during the day – they appear to have a system whereby reported content could be marked – to be crass – as more or less ‘offensive’, and then the user gets to choose what they want to see. If you’re comfortable with slurs like the n-word or the f-word being used between members of each respective community then – in theory – you could indicate so. If you weren’t comfortable with it then your community could choose to filter it out. If you wanted full access to everything no matter how vile or offensive you could choose that on your end.

There’s an obvious possible extension of this kind of approach in a distributed environment, with one (or later more) parties tagging the content, and each instance choosing what limits it placed to content visible within its bounds. I think this is a very interesting approach, and one I’d really like to see people develop more – perhaps every Mastodon instance has a plug in, subscribes to a moderation server, and pays some money towards moderation based on the number of users they look after. In return, they get a fully moderated environment, and they can tailor their settings as to what can be seen on their instance. We’re not there yet, but it’s an interesting future direction.

Another question emerged regarding users moving their content off Meta’s servers. I mentioned this possibility above – that Meta was aware that people wanted ownership of their communities, and to be able to move them to another server if they didn’t like Meta’s content moderation or monetization. An obvious question that emerged was whether or not a user who had been banned on Meta should be able to export their content and users and start up again elsewhere on their own Mastodon server. Again this appeared to be a conversation that they hadn’t quite dug into yet, but the sense I got from that was that they’d end up saying that was acceptable, but the banned user’s content would still not be visible inside Threads no matter where else they went.

Identity systems and educating the public about the Fediverse

One conversation that emerged was about the current way in which identity works in the Fediverse. Generally it’s a bit like an e-mail address. If you have your content on an instance running at example.com, and your user name is @tomcoates, then your full identity is (@)tomcoates@example.com. 

In order to make this work with the whole ecosystem, each instance (for example Threads) needs to ‘federate’ with others, or index people’s identities from different places so when someone wants to write a post that mentions me, the client can look me up and help by autocompleting my identity. Otherwise you have to know the full address of everyone you’re talking about and that’s probably beyond most of us.

This presents a few obvious questions for Threads – (a) whether they’re going to index everyone’s identity across the whole ecosystem, which could cause some problems and (b) whether or not the general public will understand the way these addresses are constructed. The indexing obviously presents some data retention issues – how do people opt in to being indexed? What are the legal implications. This stuff comes up a lot in these discussions – with search, identity and algorithmic timelines all presenting data use questions. And it’s clear that without it—and even potentially with it—this user addressing style is likely to confuse the hell out of people.

This led to various other conversations and ideas – whether or not it was Threads’ responsibility to educate people about how the system worked more generally (for example – does a user know that your content will be cached on other servers which other people may run ads against, or how the identity system functions) and whether or not other identity systems should be created. 

One option presented that has been talked about a bunch recently was using a domain name as an identity, one separate from the service you were currently using to write or consume your content. A common response to this was that this was another step into confusion and complexity. Others argued that it was impractical to try and make Threads take on the responsibility of explaining things to people and to hold their hands through the whole process, and that people would just gradually pick it up and come to understand it over time.

I want to argue exactly the opposite – that a service like Threads is very clearly going to have to explain to people how this all works. They’re going to have to find a way to make it understandable and ideally simple to the general public. And the reasons for this are twofold – regulation and bad PR.

They’re going to be legally required to write a reasonably clear Terms of Service document and Privacy policy that articulates exactly how everything works. And if the public end up not understanding what is happening, then the next time they’re hauled in front of regulators in the EU or in the US Congress they’re going to find themselves in very hot water.

I’ve written documents before explaining to people how decentralized systems work and what happens to your data and content. They are not easy things. The concepts surrounding them are tricky. It is hard to do it well. The hardest bit is to explain to people precisely how everything works (which is not that dissimilar from e-mail) but to make it sound as non-threatening as it generally really is. You have to explain to people that if you write something, it may get cached on someone else’s server and may never completely be erased from the internet, without scaring them. And explaining that the same is true of writing a post on Twitter or Facebook doesn’t generally help.

I think I’ve done a decent job of these explanatory documents in the past – I’ll post an example at some point in the near future – but it is tricky. Nonetheless, it simply has to be done.

But it’s also not such a hard thing to do. It is within the scope of human endeavor! Every single concept we take for granted today regarding the Internet and how it works was initially confusing as hell. Web addresses, cookies, user accounts, secure sign-in, privacy policies, user moderation rules, private/public accounts, using browsers, setting up e-mail, that if you send an e-mail you can’t then delete it. All of these things are things people have had to learn, and all of them started off quite hard for a member of the public to get their heads around. It was education, good interface design and clear instructions that got us to this point. And none of that will change. If Meta tomorrow set up an identity system where you chose a user name and a TLD and they said, “now you’ve made an identity that you can to login to a thousand different services” then people would start to get the hang of it. 

Still, it’s annoying work that adds friction and the general public will almost certainly start off being confused by it. Frankly, if there was any area that Meta could really help everyone with, it might be by putting its weight and presence into getting its three billion users up to speed with how the Fediverse works

Personalization and algorithms

When you sign up for Threads and follow some people, you do not by default then get dumped into a reverse chronological list of the posts they’ve written. You end up in a space that’s very similar to Instagram’s algorithmic feed. That is to say – your feed will be a mix of particularly ‘good’ (as determined by the algorithm) posts from the people you follow, mixed in with other posts from other places that the algorithm thinks you’ll like (read: engage with).

I’ve been thinking about this stuff or a while, and I have some fairly strong opinions on it. The conclusion I’ve come to is that this kind of approach is absolutely great for entertainment-style products, but actually very dangerous for news-based or informational products. TikTok is the prime example of an entertainment style product, as perhaps to a lesser extent are YouTube and Instagram. These products simply track what you watch or like and then deliver you other things that you might find delightful and interesting. A good proportion of the time the recommendations are at least fine, often they’re brilliant and 99% of the time the choices they make have little to no impact on your life or environment.

This simply isn’t true for informational or political content. In these situations the algorithm is quite capable of heavily influencing someone’s opinions and views and distorting their view of the world in a way that damages our ability to function as citizens – simply by choosing what content is put in front of them. My ‘For You’ section on Twitter is relentless in pushing right-wing messaging and commentary, even though I have no interest in it whatsoever. This is probably because I spend a bunch of time angrily debunking the worst excesses of it. The same tricks that feel honorable and positive and fun in an entertainment context (choosing things it things will get a reaction and engagement) start to feel a bit more sinister in news and political content – like the algorithm its optimized for generating fights or controversy or clickbait. It feels nasty and risks promoting division or conflict.

Meta thinks that Threads should not be fundamentally used for political commentary, but bluntly good luck enforcing or influencing that. And if it’s going to be used for those things then the presence of the algorithm is – in my opinion – a problem.

Various people spoke up about the algorithm during the meeting although it’s possibly fair to say that none of them were quite as exercised about it as I am.

I feel quite strongly that at the very least the user should be able to choose whether to use an algorithmic timeline or not, in a clear and easy to find way, without it automatically switching back after a period of time. There are some ways you can make this work in less dangerous ways – showing ‘highlights’ from your feed but giving the option to see everything for example, or having a separate algorithmic and reverse chronological feed that a user can switch between easily. And I would strongly recommend that Meta switch to one of these approaches.

In the meantime there are some very specific issues which came up during this session which are worthy of a mention – one in particular was whether or not content from third parties in the wider Fediverse should be part of this algorithmic feed. “Would they expect it?”, was one comment. “Would they be horrified by it?”, was another. 

I actually think this is a little more complicated than it initially appears – and again gets to the necessity of creating databases about people who have not opted in to Meta which I mentioned above. For Meta to be able to do recommendations of third party content means doing some kind of analysis and indexing of that content – ie. creating a profile of a user who never signed their TOS with information about them and who to recommend their content to. Quite apart from whether or not people are comfortable with that, I’m not sure what the legality of that might be from a (in particular) European GDPR perspective. I suspect if they’re careful it will be legally okay, but it still slightly sets my teeth on edge.

The final workshop

The last thing we did on the day was take a few significant groups – instance owners, Meta users and third party fediverse users – and break into work groups to answer a few questions, which I managed to rapidly capture. They were as follows:

  1. What do you think your group’s core expections are for their experiences in a federated social network environment?
  2. What do they expect of Meta vis a vis data collection and use?
  3. Are any of these expectations mutually reinforcing? Are any in tension with each other? Do they introduce any new or additional risks eg. to privacy, safety, free expression
  4. How can meta be considered a ‘good citizen in the fediverse’

I’ll be honest, these sessions were too short to be particularly useful, and I don’t know that we came up with much that was particularly interesting, but I thought it was worth posting the questions Meta were asking to give some sense of how they are at least at the moment trying to be decent citizens.

Other questions

Before I wrap up, here are a few pithy questions that people presented and the answers that Meta came back with, for the sake of completeness.

Could each person’s feed have an atom feed so people could subscribe without using the core app? 
The general sentiment in the room was effectively ‘we don’t know’, but my guess would be the answer will eventually be no. 

Would the content from third parties using Mastodon be presented as equivalent to the ‘native’ content or would it be like Apple’s blue and green iMessage bubbles?
The sentiment here was again – don’t know. Work in progress. But people obviously were keen to make sure that users understood what they were looking at.

Should Meta lead or follow in the development of the commons or the standards?
The opinion here was that it would be best for Meta to be involved but do not want to do it alone. “The real goal is not for it to be lead on all of the issues which are issues across the fediverse”.

Conclusion

My sense after this meeting was that Facebook are seriously interested in integrating Threads with the Fediverse. They do not want to crush the Fediverse. They perhaps think the future is a few large companies maintaining clients in a shared social space, where there’s also a long tail of other independent clients. This is going to weird out a bunch of people but it’s a goal I also share, so I’m okay with that. It’s worth saying that I want this world because I think it is a realpolitik alternative to one large company owning everything. I think a multi-client/company/instanced system like this would be better than now and solve a bunch of problems. But is it the utopian vision of how the world could be if we didn’t live in quite such an extractive system? Nope.

Back to Meta – I think they’re trying to engage positively, but I think that is going to be very difficult for them given the size of their userbase compared to the tiny environment they want to connect with. And I think they’re desperate for – and need – some bodies or organizations that represent the interests of the fediverse more wildly with whom they can coordinate standards and listen to desires for technical change, and can stand up for the interests of all parties. Such an organization doesn’t exist yet. Perhaps someone needs to start one.

All in all though, despite some very major misgivings here and there, overall I came out of this event much more sanguine about the way this is unfolding, a bit more optimistic about the future of the decentralized or social web, and interested to see where things go from here. I suspect we might see someone from Google or the BBC or Yahoo or Microsoft or LinkedIn make a similar move in the not so distant future. Who knows? Maybe we’ll get the interoperable shared, open social web that many of us have wanted for the last twenty plus years? Wilder things have happened. Fingers crossed?

Thanks for reading. I know it’s long and needs some editing. If you were there and want to be explicitly mentioned as an attendee, let me know – again on @tomcoates on twitter or email at tom [at] plasticbag [dot] org. It’s worth restating that these are my notes and represent what I understood from the day, and that I may have misheard or misunderstood some of the sentiments expressed. If you were there and think I left out anything important or misrepresented anything, let me know and I’ll consider a correction. Hope everyone’s having a nice day. Yours, xx Tom

Categories
Net Culture Technology

How to find me off Twitter.com

Elon Musk’s Twitter in the name of ‘Free Speech’ has just banned anyone mentioning that they have accounts on other platforms – whether that be Truth Social, Facebook, Mastodon or anywhere. 

This is an unconscionable move clearly designed to disincentivize people from leaving the platform, locking them into his ecosystem. It is petulant, childish, selfish and obnoxious, so entirely in line with our expectations of Musk’s behavior. And it is obviously completely anti-freedom of speech.

It’s also a clear and public admission that people are fleeing the platform in droves as a result of his appalling views and policies. But of course, the more he tries to trap people here, the more rapidly people will leave.

Since I’m not able to easily post my links on Twitter, I’m just going to do them on all the other places instead. So this is where I’m going:

(1) I’m moving most of my daily posting to Mastodon. The current clients are not super great and I know it’s a pain choosing a server (by default you probably want mastodon.social), but I think it’s the best option and future clients in development are far better. My address there is http://mastodon.cloud/@tomcoates – that may change in short order, but when it does, everything should transition over automatically.

(2) This is my personal blog and I haven’t been updating it very much, but I have a feeling that’s going to change relatively soon.

(3) I’m on Instagram and I use that pretty regularly, but it’s mostly just me playing with my fancy Leica or posting stories of my potting adventures, so not for everyone. https://www.instagram.com/tomcoates

(4) I’m also on Post and my username is there is @tomcoates 

If I join anything else I’ll probably update the post above, rather than make a new one.

Until a couple of years ago I was working on a product called planetary. It’s available on the App Store, and it uses a truly decentralized protocol. For various reasons I’m not on it very much and I don’t know that I can in good conscience totally recommend it, but I thought it may be interesting to some of you invested in Secure Scuttlebutt and other crypto-based social stuff.

Otherwise, let’s start this transition and go somewhere else. I’m bored to death of Musk and his poisonous drama.

Categories
Journalism Politics Social Software Technology

Trump blocked me on Twitter. But for democracy’s sake, we can’t ban him.

I was commissioned to write this Op Ed by NBC News after discussing the matter on Twitter at length. It was a fun if surprisingly hard thing to write. I never managed to get paid for it and never signed anything, so I think it’s probably okay that I republish it here. The original home on NBC News is here: Trump blocked me on Twitter.

A little over six months ago the President of the United States of America blocked me on Twitter. He or his people decided — over the course of one weekend in June — to purge those of us who had been fact-checking him online. By Monday morning, most of us were gone forever.

In a normal administration, a fairly minor micro-scandal like that might represent the high-water mark of public interest in the president’s social media life. Even in this case, there’s more to the story than perhaps meets the eye — blocking critics from official public fora could arguably be illegal — but still, I can’t imagine any previous president spending much time worrying about the effects of Twitter on their agenda.

But things have changed. Today, the censoring of President Donald Trump’s critics represents only the tiniest part of the Trump and Twitter love story — a never-ending 24/7 horror show focusing on and around a profoundly irresponsible and incompetent man’s willful and occasionally terrifying use of social media.

Let’s review: Trump — in the last year alone — has used Twitter to systematically lie to the American peopleattack the very idea of the free pressundermine public trust in America’s core institutionsunderplay racist terror actssupport alleged child molesterscall himself a geniusalienate America’s allies and perhaps worst of alltaunt the world’s most autocratic and unstable nuclear power.

It’s no wonder that so many activists now argue that Twitter has a moral responsibility to ban the president. At protests outside Twitter’s San Francisco offices earlier in January — protests that eventually triggered an anemic and half-hearted response from the company — activists argued that Trump’s appalling behavior had broken the company’s Terms of Use regarding abuse and harassment and should result in him being banned.

They also said that Twitter’s founder Jack Dorsey — by creating a space where Trump could circumvent normal media checks and balances — had directly contributed to the president’s rise to power. Enough is enough, they argued. Ban this man.

I have a lot of sympathy with this argument. I also know some of these activists personally and they are honorable and decent people. But ultimately I believe Twitter must fight to keep Trump on the platform.

For good or ill, Twitter is one of the closest things we have today to a de facto “public space” on the internet. I believe we need such a space. And I believe over the last couple of years, under extraordinary (if deserved) pressure, Twitter has just started to really understand the full range of responsibilities that occupying such a role entails.

One of these responsibilities is to provide a space for the political discourse of a country to play itself out. These are the spaces we now use to debate the issues, to campaign and — now — even to discuss and announce policy. Ideally they wouldn’t be spaces owned by for-profit corporations, but truly public places with rights and responsibilities defined and protected by law. But the U.S. government has shown no inclination or ability to fund or build or run such places, so instead we are where we are.

And where we are is in a country where almost half of the electorate voted for Trump. He did not organize a military coup. It wasn’t a massive administrative error that secured him the job. It was, as much as some people may dispute or dislike it, the will of the people. And until such a time that he’s removed from office, if Twitter is to remain the de facto public space we all need, the will of the people matters.

I’m not going to pretend there isn’t realpolitik in play here too. Let’s face it: Banning the president from Twitter would not remove his platform, he’d simply move to Snapchat, or Facebook or Ello. And if he were banned, the partisan outcry over the decision would probably rend Twitter in half in the process, potentially killing the product and the company in the process. There are no victories there.

Because in the end, the only victories can come from the same processes that got us here. We need to take responsibility as an electorate. If we want him to stop debasing the presidency on Twitter, we need to remove him from the presidency, not remove him from Twitter. We need to support our courts in the fair implementation of the law. And we need to hold our elected representatives to account as they attempt — in turn — to keep Trump from going off the rails.

Meanwhile, there is something we can ask of Twitter. We can ask them to be clear about how they see their role in the world. We need to know what they believe in; what they stand for. We need them to demonstrate that they fully understand they’re not simply a neutral communications mechanism. Today’s Twitter is a place where business happens, elections happen, government happens — and with the arrival of Russia onto the scene — international tensions play out. We need Twitter to show us they understand this and that they’re up to that challenge.

And perhaps we can ask them one more tiny thing — to review their policies on politicians blocking or banning users engaged in legitimate, non-abusive political debate. Twitter’s own statement stood up for “necessary discussion around [politicians’] words and actions, but we can’t have that discussion if those politicians shut us down. And in this post-truth world, we need all the help we can get.

Tom Coates is an entrepreneur and technologist who has developed software products for the BBC, Time Out, Yahoo, Nokia and Jawbone among others. Over the last 20 years he’s written and spoken extensively about tech culture, social platforms, location and the Internet of Things and his work has been featured on the BBC, The Guardian, New York Times, MIT Technology Review and in the Daily Mail. His most recent project was the smart home software company Thington, which was acquired last year by eero inc. 

Categories
Design Internet of Things Talks Technology

The Shape of Things

This is a transcript of a talk I gave at Webstock in New Zealand in February 2016, lightly edited to remove many terrible jokes. For more information about Webstock, New Zealand, the artwork in this piece and more more, skip to the bottom of the page to the unnecessarily long transcript. It’s also a repost of an article I posted to Medium. See the talk in its original context.


What I’m going to be doing today:

Today I’m going to be talking about the thinking we’ve been doing at Thington about the right and wrong ways to interact with a world of connected objects, and some of the problems we’ve been trying to solve.

In particular I want to talk about the relationship we’re starting to build between physical network-connected objects and some kind of software or service layer that sits alongside them, normally interacted with via a mobile phone.

And I’m going to talk a bit about how there’s a push in the design community to find a different model, dissolving the top layer here into the object itself through (a) tangible, physical computing, or through (b) metaphors of enchantment or magic:

I’m going to try and argue that both of these models are kind of wrong! And I’m going to be chatting about a few ways that I think we could and should be a bit nicer to the software or service layer (with a nice long digression about tangible computing on the way).

This is, by all accounts, a pretty deep, weird and nerdy talk, through which I hope to expose to you some of the insane depths of computer history and the weird arguments designers have.

But first a little history…

A brief history of computing in the Twentieth Century:

We’re going to start with Thomas Watson — the gentleman founder of IBM — and a statement that he is alleged to have made in 1943 that sounds crazy and entertaining to modern ears. That statement is:

“I think there is a world market for maybe five computers”

There’s actually very little evidence he made this statement at all, but at the time it wasn’t a particularly unusual statement to make. For example, Charles Darwin’s grandson — who was slightly unfortunately also called Charles Darwin — said in 1946 of the UK:

“It is very possible that … one machine would suffice to solve all the problems … demanded of it from the whole country”

And then there’s this chap, early computer pioneer and telepathic supervillain Howard Aiken, who said:

People genuinely didn’t think there were going to be many computers in the world! Even after twenty years — maybe because of twenty years — in the tech industry I find this a super weird thought. I find this a really hard idea to get my head around.

So how did this match up with the actual reality? This following picture is me in the late seventies in my favorite Disney Winnie the Pooh t-shirt, in Norwich in the UK, not looking very cool. This is about thirty-five years after Thomas Watson from IBM’s statement, coincidentally roughly halfway between that statement and today:

In terms of computing, where are we? From the four or five computers that Thomas Watson thought we’d have, we’re already up to the massive 50,000 units of computers sold each year. That’s quite a shift!

Skipping forward another fifteen years or so, here I am again:

Here I’ve finished primary and secondary school, and I’ve gone to University and I’m starting grad school and I’ve popped over to the US to have my photo taken on top of the Empire State Building with my tongue out. At this point in the world there are in active use something like 150 million computers.

Here are pictures of me in the early 2000s (one billion computers have ever been sold) and the mid-2000s (two billion computers have ever been sold).

And then of course this happened:

I think we all forget how quickly things can change, but I think it’s fair to say that the era of the modern smart-phone starts with the iPhone, and it’s really important to remember that only launched a little under nine years ago. This by the way, is the very first advert for the iPhone which essentially replaced single use telephones with general purpose computers connected to the phone network.

Three years after the iPhone launched — so about six years ago now — in addition to all of the desktop and laptop computers we were buying, we were also buying 150 million smart phones a year.

Five years later — 2016 — and it’s projected that 1.6 billion smartphones will be sold. In one single year, one smart phone will be bought for every five people on the planet.


But what happens next? A world of connected objects.

Now the reason I’ve taken you through this little adventure is to just remind you that within a human lifetime, we’ve gone from essentially zero computers sold per year to billions. It’s been a period of an extraordinary increase in the availability of computation — with processors shrinking and becoming more powerful every day. And not only has it been growing at an extraordinary rate, that rate itself is accelerating. The last decade has seen a massive expanse in available computation and it shows no sign of slowing down. We can expect a world of hundreds of billions — trillions — of computers distributed around the world around us within a few years — embedded absolutely everywhere they can make even the slightest incremental improvement.

I’m talking of course about the Internet of Things, and this is where I make my first grandiose assertion of the day:

It’s a time of tremendous change. After years of design experiments and academic discussion, the cost and availability of components and the ready availability of smart phone interfaces means that the Internet of Things is finally rapidly approaching.

In fact, I’ll go further and say that within a decade almost all new electrical appliances and devices that we buy for the home will have some kind of network component — to say nothing of our offices or public spaces. Quite seriously, the world of tomorrow is dripping in objects that belch out information or can take commands, or both…

But don’t take my word for it. This is Samsung’s CEO at last year’s Consumer Electronics Show in Las Vegas, where he made the Internet of Things their major focus:

“By 2017, all Samsung televisions will be IoT devices, and in five years, all Samsung hardware will be IoT devices”

By the way, CES is an amazing event too. After years of it being something that internet people didn’t really attend, it feels like that’s finally changing as software and computation starts moving into devices.

When I went last year, there were smart dishwashers, heaters, smart air conditioning units and humidifiers, smart lighting, smart garage door openers, things to open curtains, check if your house was on fire, smart ovens and kitchen scales, smart vacuum cleaners and smart security systems. If it could take a battery or plug into the mains there was a smart version of it… And they were being made by companies like Samsung, Polaroid, Canon, Panasonic, Quirky, Sony, Belkin, Parrot, Honeywell and many others.

But how do people access the power of these devices?

Honestly, however good the hardware was at CES and however much of it there was, often the the benefits that thetechnology seemed to bring people just did not seem to be as good as I might have hoped. The power of the internet was just not present in these devices — the networked parts of the product were effectively little more than app-based remote controls.

To me, it was still clear that they were going to be able to offer tremendous power to us all to control and understand the world around us, but there was little sense of how a normal person might harness or grab that power in comprehensible ways.

Now some of you might be even more suspicious and say that there’s little or no power to be discovered in a world of connected devices. I think that’s wrong, but I also think it’s an understandable statement. Some of these devices have been crazy and lurid and wasteful in their use of technology and bring no obvious benefits to their users.

But today, I’m going to focus on those devices that (when enhanced with the internet) get better. This statement from my friend Matt Rolandson from Ammunition Group in San Francisco sums up this category for me:

“You should use the network to amplify a tool’s core purpose, not to be another web browser or Twitter client”

I really think this is one of the most bluntly useful and apparently obvious things that anyone has said about the Internet of Things as it manifests in devices and appliances. The Internet can and should be used to amplify a devices core purpose and if it does so, it makes that device better and more useful.

But surely there are better ways for us to amplify the purpose of things than just giving them a remote control?

There now follows a brief guide to Tangible Computing — and thanks to Doctor Who and Rick & Morty for the examples I’ve used here.

Merging the physical and the digital

Okay — so this is the pattern that I mentioned earlier — essentially there’s an object — here an oven — and it comes with an app that runs on a mobile phone. The app is essentially a remote control for the main object, one that potentially has a few rule making components with it that make it a little more interesting and useful.

This is a model that actually enhances the object it’s attached to — it makes it easier to control or check up upon from a distance, but it seems a bit simplistic and on the nose. Can this really be the extent of the future we’re looking for? And is the reason it feels a bit dull an interaction problem?

One direction that designers have traditionally been very keen to explore is dissolving those two parts together — merging the service layer and physical objects to make something seamless and more powerful, that a user might interact with in both the ‘data’ environment and the ‘physical’ environment at the same time. Essentially the goal here is to break down the distinction between the two parts of the ‘thing’ (actual vs virtual) to make something new and hybridised.

There are so many people making these arguments in the (perhaps more cerebral) parts of the design community that it sometimes seems inevitable that this is the direction that we should be moving in. Let me give you some examples:

This is a quote from Josh Clark (from Connected // Disconnected), who has been talking around IoT on the design side for a while now and is a very sharp guy:

“The potential of the internet of things is to improve on what mobile does so well. Instead of availability at the point of inspiration, IoT lets us shift to interaction with the point of inspiration. Add sensors and smarts to an object or place, and you no longer have to pull out your phone for a digital interaction.”

And this is a quote from Matt Webb in the UK from a piece called ‘Waving, Not Designing’ that he wrote a few years back:

Why use just your fingers to select what’s on a display when you can use your whole body? It’s often easier, and makes more sense. Like, when you use a hammer, you don’t key into system to say “hit at point X with force F” and then stand back and let it happen, you just pick up the hammer and hit with it, using your body to judge strength and your eyes to judge position.

I could genuinely list a hundred other people making this argument. And they’re some of my favourite people in the design community too — people who are looking around, reaching for something truly new and interesting and more intuitive.

This is literally a screenshot from an argument I had on Twitter with a super sharp friend of mine, who is clearly pining for something a bit beyond the model of “phone + thing”.

Designers are looking for a new natural vocabulary for the next generation of devices, and they’re looking towards embodied interactions and tangible computing.


A brief guide to Tangible Computing

Much of this thinking is inspired by the ground-breaking work of people like Paul Dourish, David Rose, Durrell Bishop, Natalie Jeremijenko and Hiroshii Ishii.

Now I have to apologise here because for brevity I’m going to have to wildly over-simplify their positions (you should go and read and research their work online at least — it’s great stuff) but essentially they want to blur and even dissolve the distinction between the digital and the physical. They think rather than have a differentiated service layer, the magical intelligence should merge with the physical object. And that, in the doing of this, they believe simpler, clearer, more powerful, magical objects emerge.

Their argument is fundamentally that the world of screens and icons is too abstracted and separate from the world around us and the ways in which human beings understand that world.

Let me give you an example — David Rose in his book “Enchanted Objects” talks about four visions of the future. The most awful one he describes is called, “Terminal World”.

“It is years into the future. All the wonderful everyday objects we once treasured have disappeared, gobbled up by an unstoppable interface: a slim slab of black glass. Books, calculators, clocks, compasses, maps, musical instruments, pencils and paintbrushes, all are gone. The artifacts, tools, toys and appliances we love and rely on today have converged into this slice of shiny glass, its face filled with tiny, inscrutable icons that now define and control our lives…”

Now unsurprisingly, David doesn’t want this world to happen (or for phones to eat his children) so he presents an alternative view that he called “Enchanted Objects”. He describes it as ‘technology that atomizes, combining with the objects that make up the very fabric of daily living’ and the examples he provides are really lovely.

Glowcaps helps people remember to take their pills

One particular example he worked on himself is the Glowcaps system, which beeps and flashes with increasing urgency if you forget to take your pills on any given day. And he talks about many others, including Nest Thermostats that predict your temperature needs, umbrellas with lights upon them that signal up when it’s going to rain that day and many more.

These are genuinely useful and interesting things and there genuinely are more of them every day coming into the world. The Glowcaps alone have a huge impact on people whose drug regimens have to be strictly adhered to.

The metaphor here, as I’ve said, is ‘enchantment’ — magical interactions — bringing the intelligence into the object itself as you would with an ancient sword, rather than believing in the presence of a separate, service layer.

Diamond Pickaxe + Enchanted Book = Enchanted Pickaxe

Leaping back in time quite a long way for a moment for illustrative purposes, here are two classic examples from the early nineties of the blurring of the physical and the digital — bringing the virtual representation and the real object so close that they become one definite thing.

On the left we have Durrell Bishop’s answering machine. This pumps out a little marble when you have a message, and you place the marble on a sensor tray to play it back. Natalie Jeremijenko’s dangling string simply indicates the amount of network traffic in a space by twitching a string in a room, giving people an ambient awareness of activity.

I bring these up because they are classics of the field — almost foundations of the field — of tangible computing and were first to articulate some of these goals that we’ve talked about so far..

Paul Dourish, in fact, went a step further in his seminal book ‘Where the Action is’ around 2001, suggesting not just that things would be better if the physical and the digital were closer together but that such a path was effectively inevitable and natural.

He framing our interactions with technology as a series of approaches that build one upon the other, each employing a skillset that more closely reflects how human beings understand and instinctively interact with the world.

Paul Dourish’s “The Stages of Human Experience of Interaction with Computers”

And it certainly does seem like there continues to be a lot of ways in which more tangible interactions with enchanted objects could provide a lot of power in the world. It’s clear we’ll see a lot more of this kind of approach — the focusing on the invisibility of the technology, dissolving in the use of the object. It promises a certain seamlessness of interaction.


Some problems with merging the virtual and the actual

But is it the ultimate answer to how we interact with a world of connected objects? There’s a desire by people using these guys as inspiration to try and make every object self-explanatory, self-evident, complete and seamless and separate from other things. And that seems like a flawed enterprise to me and it seems to miss where quite a lot of the power of connected objects might be… That is, in the connections.

In the first place, I think there’s a bit of a category mistake going on here. For Hiroshi Ishii and Paul Dourish, for example, the work they’re doing is more concerned with using physical interfaces to manipulate data, rather than bringing computation into devices. Their focus seems generally in making the manipulation of digital objects more intuitive by bringing it into the physical, a space that we have dedicated millions of years of evolution to understanding intuitively.

The internet of things, however, is much more about enhancing the physical with the digital, making the objects make more sense at a distance, or drawing out information from them and bringing it into a virtual space where we can do stuff with it.

In some ways, you might argue that the fact that the two merge the physical and the digital is a coincidence — and that in all the ways that count, they are actually opposites of one another.

I’d also add that one of the thing I think contemporary designers miss is that these thinkers were very focused on the environment surrounding the object and the abstract information about who owns it, who can use it, what information the object needs in order to be able to do its job most effectively.

Dourish is very focused on the environment around the things, Rose very focused in the services around the objects. I think it’s a mistake to think that their focus on better objects means less focus on better service layers.

Another example that complicates this tangible vision — in my home, I have various smart lights connected to the Internet but it doesn’t really make sense to me to think of them individually — they’re part of a larger system which is ‘My environment’. What is it specifically that I connect with or touch or interact with to make them act in the world? The objects are definitely acting independently, but it feels like there’s something that connects them.

Again, it feels like something that isn’t situated in the object, it feels like there’s something between them. My intuition is that I’m communicating with or manipulating something beyond the level of an individual object, and it feels like that’s the intuition that in this context we should build interfaces around. Perhaps then the power simply doesn’t come from dragging the network down into the physical thing, but with embracing the network and the object as complementary but separate parts of the same system…

In fact it’s this problem of what’s most intuitive that gives me most pause for tangible computing generally. The assumption from many of these thinkers is that making an interface that’s physical makes it inherently more intuitive.

But I don’t buy that physical affordances alone will make it immediately obvious what a smart connected object is for. Sure, you pick up a hammer and you immediately want to hit something (or maybe that’s just me) — but is that true of a smart hammer?

That seems to me to also be dubious — for every good product that makes more sense when embodied or made tangible, it seems another is likely to pick up some strange magical interaction metaphors that are less intuitive, or even counter intuitive. It’s quite possible that in taking something that is ‘natively’ digital or abstracted and merging it with something else with physical affordances, we create a thing at war with itself. Not more intuitive at all, but just much more confusing.

Are we making things that are effortless, or are we simply creating a whole new vocabulary of interactions that people have to get their heads and hands around?

Our cousins in computer engineering talk about General Purpose Computing — whereas as designers we’re often tempted by the quest to find the ultimately specific interface for the thing in front of us. But each slightly different interface creates an extra cognitive load that when multiplied across every object in the world may be wildly less intuitive than a General Purpose Interface on a phone, or smart watch or computer whose abstracted rules we learn once and can then apply everywhere.

In this quick diagram I knocked up, the green stuff is the thing that is immediately intuitive and doesn’t require learning. The red stuff are the bits that you have to learn to use. My argument here is perhaps a slightly harder to learn ‘General Purpose’ UI might have less cognitive load than a whole bunch of nearly intuitive devices where there isn’t any transferable knowledge.

So if the solution isn’t merging the physical and the digital, what is it?

Maybe we could do more with the service layer instead?

Towards a stronger service layer?

Personally I think the solution of how the physical and the digital should interact is not to bring them closer, but instead try make the relationship between the two clearer and then push the power of the service layer far beyond where it is at the moment. And I think when you actually look at the problems that confront people when they use IOT devices, you end up essentially defining the properties of what that service layer should be.

In a moment I’m going to tell you a bit about what I think that picture looks like, but first — if you’re a designer who is in any way uncomfortable with this idea of the point of interaction and the device being separated from one another — I have a quick example for you from history which might help you relax.

This is the evolution of the light switch. Lighting started with controls directly next to the kerosene, oil or electricity light and gradually moved away from the object itself to the places they made most convenient sense — by the door that you walk in through.

You shouldn’t doubt it took them a while to get there — I love the one with the cord in the middle that uses a sort of pulley system to put the interaction where it makes most sense — but in the end we all decided that we understood that the right place for a light control is just where you want to turn on the light. The light is the thing. The service layer is the switch. And the service layer sits wherever it makes most sense for the person, with the relationship between the two clear and simple.

Here’s a more up-to-date example. Zipcar has cars parked in garages and parking structures across the World. And you can book them online or from the app and open the doors and drive off with them at any time with a simple RFID card.

But the hardware here is trivial — it’s just an RFID reader and a couple of switches, allowing the engine to start and the doors to unlock.

It’s in the service layer that the value of Zipcar truly lies — you probably book the car from home, so you’ll probably book it via your phone. And then you’ll use all those brilliant features of the internet that cars don’t naturally have — an understanding of identity, payment and a sense of location. It’s from that interplay a beautiful and powerful service is born that makes thousands of cars in thousands of locations yours to spin up as a software engineer might spin up an EC2 instance.


So what is the ideal service layer for the Internet of Things?

Here are the six things that I currently think are the core features of an enhanced service layer for general smart things.

Number one is nice and simple — the ideal service layer gives you control. It should give you the ability to control an object locally (even though it may be easier to do it through a physical interface) as well as from a distance.’

This is so obvious I’m surprised I have to mention it, except that advocates of embodied interaction always seem to miss that it’s actually a core attribute of a smart device that you don’t need to be physically present to control it or find out information from it. The exciting part of the Internet of Things is the Internet! And the Internet has been about collapsing distance and making the world accessible wherever you are. It’s no different with the Internet of Things.

Number two is about how a service layer lives with you over time. The ideal service layer supports you from initial set-up to the day you decide to recycle it.

This is one of the things I think is most bizarrely missed out on in most IoT products. Owning a piece of hardware is a relationship with a beginning, middle and end. You start off researching something to buy, you choose it, install it, use it, try and set it up to meet your needs, you buy supplies for it, you clean it, occasionally it breaks down and you throw it away, or you get it serviced and fixed. Eventually you decide to upgrade it.

Having a service layer transforms a thing from something a manufacturer sells to something that forms an ongoing relationship between manufacturer and consumer. There’s so much potential there it’s startling.

Number three is a huge one for me and again brings in some of those features that we saw with Zipcar (and interestingly come naturally to light switches). The ideal service layer understands that the device will be used by multiple people.

Again, these are things that if you look at almost other part of the Internet are obvious and baked in. Loads of services build on the Internet have concepts of identity. You can log in as someone and get access to various features. Different people can have different permissions. But for most IOT devices this is still completely absent. (This is a subject I’ve written about before in a piece about Thington: Why people are the most important part of a world of smart things…)

A quick example if you buy a nest thermostat first you install it in your home, then you create an account so that you and the thermostat are karmically connected. Then for every other person who lives in your house, or may come and visit and who you think might have to have some control over the temperature while they’re there, you simply sign in as you on their phones too.

This, I might suggest, is crazy. It makes it effectively impossible for it to react differently to different members of your family! It makes it impossible to know why a room is the temperature it is.

It also makes it possible for someone to come and stay at your house and then once they’ve left your house somehow continue to control the temperature in your house at long distance with you having no way to stop them! This has happened to me and it simply shouldn’t be possible. We should know better. This is easy to fix!

Number four is where all the promise of the Internet of Things lives and yet sometimes feels like the farthest away from coming into reality. The ideal service layer is able to work easily with all the things you have.

A smart light switch is great, but even better if it can coordinate with motion sensors in the house and with the geofences triggered by your phone to turn off precisely when you want them to. A sprinkler system works particularly well if it knows not to turn on when the windows are open, or when it has recently rained, or if you’re just about to walk through the garden.

But even if you have number four, then you still need number five for this power to be even trivially available for people. The ideal service layer does not expect you to become a programmer.

To create the kinds of coordinated responses I just talked about, someone has to somehow encode the expectations and the relationships and string them together. And at the moment there really aren’t any good ways to do this.

One way to stop you having to become a programmer is to make all the decisions automatically for you. This is the way that Nest attempts to do things — it’s just supposed to observe how you live your life and intuit what you want to do next.

I’ve interviewed dozens of people who have the nest and with a few exceptions they’ve all turned its magical learning features off. It just wasn’t doing the right things at the right time. I should add that they all loved their nests and found the ability to warm up the house on their way home really really useful. But the predictive things were making assumptions about their activities that were not immediately comprehensible by their users. And when the learning features were on the devices felt inscrutable to them, confusing and alien.

Nest thinks you know what the problem is just as well as I do.

But if Nest’s interfaces aren’t perfect, at least they’re vastly superior to the other end of the spectrum:

Yahoo Pipes

This is an example of the UI from Yahoo Pipes, but honestly it’s the visual equivalent of what a lot of programmers I know are doing in their homes with IoT devices. It’s pretty clear this can’t be the direction.

It’s possible to simplify this kind of interaction with services like IFTTT (If This Then That), which let’s you set one ‘trigger’ and then ‘one response’ in a pretty simple way. But while it’s simpler to assemble, to make any complex situation you end up making lots of simple rules instead of one complex one. It doesn’t really make things much better. And that’s because even the simplest requests a person might make actually end up being much more complex than the first appear. For example, people say they want this:

“When I get home, turn on some lights”

But when you actually dig into what they actually want — when you take into account the various devices, people and contexts that impact how you’d like your home — you end up with something a bit more like this:

I genuinely think giving people all the power of a complex rules system, without bombarding them with UI complexity, is the hardest problem in the Internet of Things at the moment and the one most deserving of extraordinary mental effort trying to fix. All the power of these devices hides between interfaces and metaphors that are totally incomprehensible to normal people.

Finally, we get to number six. The ideal service layer communicates clearly and politely in ways that are timely and familiar. And this is I think super important, because at the moment a whole bunch of devices that we use are pretty much totally inscrutable and we don’t know why they did the things they did. And when they do decide to tell us things they do so by aggressively beeping or sending us notifications by the truckload. Finding the model that makes these communications humane and polite is one of the other largest challenges we face.

Personally I think the answers here lie in the work we’ve been doing to make communication between people comprehendible — our social streams might be a great metaphor for a world of communicative devices. But more on that shortly. In the meantime, here are my six principles again:

These are the principles that we’ve been working with when we’re building Thington — but I think they’re equally applicable to almost any software layer for a connected object you can imagine.

They are a set of ideas that I think represent a service layer way beyond the simple idea of a remote control — a service built with them could follow you across devices, across contexts, across the world. I really think this is the way that we should be working, the direction we should be pushing in.


Conclusion

So I’ve talked at length through this piece about the core directions I see in front of us now in trying to make a world of connected devices comprehendible to normal people. One is a common refrain among talented designers — that interaction should be embedded more into the physicality of the things themselves. The other is my own position, that this world at scale only really makes sense — can only really be intuitive — if we accept that a service layer exists, has to exist, is useful and important. I’ve also argued that we should push the service layer forward away from being something bland and slight like a universal remote control towards something deeper and more interesting.

With Thington in particular we’re experimenting with a couple of metaphors that I think really embody these principles and push them further in interesting ways.

Firstly we’re treating the way your objects communicate with you the same way that Facebook treats your friends communicate with you — with human readable chatty, social media-like streams of information.

And secondly we’re trying to replicate the feeling of a butler or assistant suggesting things he can do to make your life better. In doing so we’re trying as much as possible to take the complex rule-making systems away from the general user.

Obviously, I’d love it if you went to thington.com and had more in depth look at what we’ve chosen to do, but if you don’t have a chance to do that, here’s a way of representing what we’re doing that is pretty fun (even if it does make it look super ridiculous).

Thington as a home for conversational devices and for the Thington concierge (or robotic butler)

These are just our attempts to live up to the principles we’ve put together and build an experience that takes the service layer way beyond what exists at the moment.

This may resonate with you, or it may not. I hope it does. But even if you don’t agree with me on the specifics — even if you think that tangible interactions and embodied interaction are the future of every device on the planet — the one thing I really need you to believe and take away with you is that this world of connected devices one way or another is coming.

Every day more devices, appliances, sensors and actuators, homes and cities are coming online in one way or another, and this is going to have a transformative effect on the world.

I personally believe that the company that creates the service layer for the Internet of Things could be as significant, powerful and large as Amazon, Google or Facebook — only they’ll not be the way you interact with your friends, but with the entire physical world that surrounded you. We’re on the brink of a new service layer for the physical world that operates at a truly planetary scale.

And the design patterns and interactions for this world are being formed right now, by people just like us. And if we don’t get involved and design and think about the complexities of the world, then other people will, and when they do they’ll encode in them ethics, belief systems, views on privacy and intrusiveness, a sense of the role of network in the life of the individual that may be very different from the world we’d like to live in.

This has been a long piece, so it might be difficult to stretch your memory right back to the beginning, but if you remember, I referred to this alleged misquote by Thomas Watson, the founder of IBM.

“I think there is a world market for maybe five computers”

When I read his line, I’m always reminded of Clay Shirky’s response, referring to the Internet.

“Thomas Watson predicted that the world would not ever need more than five computers. We now know that he overestimated that by four” Clay Shirky

To which I would add only add that every day, more and more, this one computer in the world is the world in which we live.

It’s massive, wild, distributed and it’s starting to break free of the browser and the app and permeate every aspect of our homes, offices and public spaces.

So this is the time to get involved, to explore this space and find better patterns, better interactions, better models of how the future will work. This is the moment where we as designers can have the most impact, helping to define a User Experience, an ethical, powerful, transformative UX at quite literally a planetary scale.

The world of tomorrow could be transformed for the better if we work to make it so, and I believe very strongly we have it in us to make it truly extraordinary. And that’s all I have.

Postscript:

If you’re interested in trying out Thington, go to https://thington.com to find out more and download the app. If you have any questions or comments, ping us on our Twitter account @thingtonhq or e-mail us at hi@thington.com and we’ll do everything we can to help you out. There’s a list of Frequently-Asked Questions on the Thington website too: Frequently Asked Questions.

If you’re a manufacturer or potential partner and you’re interested in Thington integration or want to find out more about what we’re doing, then e-mail us at hi@thington.com.

If you’re a member of the press and would like to talk to us then e-mail us at press@thington.com — there are also some resources for you available at https://thington.com/press

As I said at the beginning of this article, this talk was given originally at Webstock in New Zealand in February 2016. In this version I’ve removed a few of the jokes that only make sense if you know Webstock orNew Zealand very well. Having said that, if you haven’t been to New Zealand or to Webstock then you are really missing out and you should make an extra special effort to do so. It is my favourite event on the planet, the country is stunningly beautiful and the people are completely amazing. As always, thanks enormously to Tash, Deb, Mike, Ben and the Webstock audience for being incredibly welcoming and brilliant in every way.

The artwork in this piece was mostly drawn by Tom Coates in Adobe Illustrator, but the visual style of the whole thing comes from a beautiful piece of illustration by Chris Martin for Wired UK about setting up a smart home and my own efforts with the House of Coates twitter account. I loved the piece so much that I found a way to contact him online and I have two copies of it printed and framed — one in the office, one at home. My own efforts in this article are a bad pastiche of his extraordinary work, and I hope are received with the spirit they were created — as a statement of enormous respect to his creativity and a love of his style.

The typeface in the illustrations to this piece was made by Tom Coates in a weird little app for the iPad called iFontMaker. It is — as you might expect — based on a slightly stylized version of my handwriting. If you want to get your hands on it, please reconsider! There are much better typefaces in the world with a similar aesthetic. If I still can’t persuade you otherwise, I’m still working on all the special characters, and when I’ve done enough of them I’ll think about putting it out in public.

This material is in part based upon work supported by the National Science Foundation under Grant Number (1621491). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

Categories
Design Technology

An extremely quick and rough attempt to explain design and designers to people from other disciplines… V0.2

Warning: This is rough as a bucket of spanners, but I’m trying to summarize a whole bunch of stuff quickly. Feedback is appreciated. Work in progress etc.

Watching someone collide with some of the various ways that designers tend to describe themselves and how their job titles have shifted and changed over time, I tried to write up a quick summary of how I understand design practice, particularly focused on people working with products that have digital components. I’d love to get any feedback or thoughts that any of you might have — feel free to add a comment on the side or query stuff, and at some point in the future I may attempt to write it up in a more coherent form.

A particular problem I’ve noticed is an absolute brain fart that I’ve sometimes had from people who can’t get past that designers are fundamentally about making things ‘pretty’.

Well for a start — yes we are. A huge part of design is working out how to communicate a product, sell a product, make the experience of using a product evocative, emotionally enriching and pleasant. But there are other skills that are lumped under ‘design’ as a discipline and any individual designer you meet will have strengths in one area or another to a greater or lesser degree.

This piece is just an attempt to summarize some of the things that people who operate under the title of ‘designer’ might focus on, or be good at. Some designers are very flexible across the spectrum, others focus in on one area or another and are deeper in that one as a result.

At a first approximation in a primarily digital environment these are the kinds of things that designers are focused on:

Look & Feel / Visual design:

Does it look and feel appealing / professional etc?
Do you have an appropriate emotional reaction to the thing that makes you want to use it more?

Communication design
(in the context of internet products):

Does a user know what the thing you’re making is for?
Can they tell what kind of thing it is from looking at it?
Does it reveal how it’s supposed to fit into your life?
Is the language used clear and unambiguous?

Interaction Design
(and a bit of information architecture):

Can you tell how to use it, does it make sense?
Does it work like you expect it to?
Does it make it easier to do complicated things?
Do you get a sense of how everything fits together?
Would you know where to find a particular feature you wanted?

Experience Design, Audience Research & Invention:

What are the moments in the lifecycle of the thing that you’re making that are opportunities for it to explain itself or communicate with users?
Does an existing product match with the needs you have in your actual life?
Do users have needs that are unmet by existing products and services that you could meet with a new idea.
Is there a way to communicate an idea or a concept that reveals something about the world through an object or a project.


This is a bit of an arbitrary arrangement roughly arranged to go from details of execution through to the concept —or from the concrete to the abstract.

In this arc, the ones nearer the top are focused less on the guts of the project and more about improving and enhancing it. The ones at the bottom are more about the very idea at the heart of the thing, that will be expressed in the crafts higher up the ‘stack’…

Generally the more clarity you have about things nearer the bottom, then the easier and more focused the thing will be as you go up the stack towards the visual layer.

It’s easy to say that the visual layer, therefore, is less important — ‘just coloring in’ — but I’d protest about that. The coloring in is the thing that people respond to first, the first thing they see, the first experience they have of your product, and something that can frame and color every experience they see with the product. Plus it’s bloody difficult and hard to do well. A designer who can make an evocative product but can’t think through the deeper issues systematically is no worse than a great structural thinker who can’t communicate their ideas well. People have strengths in different areas, and you hire to try and balance them.

would say though that beautiful design that expresses the essence of something probably has to understand that essence, so some knowledge of what’s going on in the guts is fundamental. But that doesn’t necessarily mean those people have to be working intensively on the core idea. Sometimes they just help you refine your thoughts and digest them into a simple narrative that they can express visually.

And yes, you’ll notice lots of overlap with other disciplines, particularly ‘Marketing’ and ‘Product’. Unsurprisingly, if you’re going to build something, then you can’t just split up decisions between the disciplines. There just IS overlap between disciplines, and if there isn’t, all it really means is that design or marketing or product is taking on responsibilities that are normally (but not always necessarily) shared or split.


While these are the things that designers are there to do, there are loads of different techniques and strategies for how they can do them from card sorting to interviews to paper prototypes to personas to wild drug-fueled ecstatic dance binges, through to wireframing, mood boards and a billion other things. Quite a lot of design conferences are about the specific techniques people use in different areas of the design stack and how useful they are, ways to be faster at the process, or more thorough on the process, or how to integrate design practice and engineering or product in various effective ways. They’re also about how the various shifts in context that happen as devices become more powerful, more flexible, change shape and interaction patterns effect the practice and the techniques you’d choose to employ. etc. etc.

The titles that people use are, at some level, attempts to articulate what part of the design spectrum the person feels their strength is, or what particular design need a company has. Do you want a designer who can make your product clearer and simpler to use? Then you’re looking for an interaction designer above everything else. If you want a designer who can figure out how to make a billion dollar product fit better into people’s lives to assert your dominance in an area, then you’re looking for someone who will do interviews, research, ethnographic studies and usability testing. If you want your product to resonate a bit more with people emotionally, and for people to get a bit more of the feel of a relationship with it, then a visual or communications designer is probably where you should be starting.

One thing designers do a lot which I think confuses other disciplines is talk about what design is. I think this is mainly because almost everyone in tech products spends a bit of time doing some level of design work to a greater or lesser degree. If you build an interface for something, or a web page that connects to other web pages in a way you’re doing design work, in the same way that if you learn a bit of HTML you’re doing some software engineering. There’s a job of work for designers in continually re-explaining and re-marketing what they do that’s different from — and ideally more professional than —the work that other disciplines do in these areas. This article probably falls under that remit a bit.

After all there are engineers (mostly in tech start-ups and large engineer-led companies) who consider some of the things I’ve listed above as part of their discipline and only let the designers come in to do the work of polishing the front-end to a singular shine. On the other hand there are designers (mostly in agencies) who don’t let engineers do anything other than implement their work and wouldn’t listen to them in any area that has a connection to design. There are product people who are focused on marketing and metrics and leave all this other work completely to designers, and there are others who are much more involved, or own the product definition stuff completely and contribute to the way the design unfolds. And all of these approaches are perfectly legitimate ways to work and build things, if you’ve hired the right people for the job you want done, and if across your company you’re covering the range of skills that you need to make a product well.

A particular weakness of mine is visual design and branding. I’m far from awful at it, but I never feel like the visual layer of a product I design represents the thing I’m making as clearly and effectively as I’d like. There are people who are much, much better at that stuff than me, and I like to work with them very much indeed. The intelligence and skill a good graphic designer or brand expert can bring to a problem is a much undervalued skill in the tech industry.

My particular sense is that different disciplines lead different kinds of companies but that the best products emerge out of an interplay between business (looking for money-making opportunities), engineering (representing the space of all possible-to-make products) and design (whose focus is on translating between the code/business and members of the public).

In the end for me, design work is fundamentally focused on the form something takes and how that connects and communicates to the people who are going to use or engage with that thing. But form isn’t separate from function, or business or material science or computing or any of the other things that influence what the product or object or made thing is and how it works in its wider context and so you can and should expect a practicing designer to have some varying interest or engagement with those areas if they’re going to do their job effectively.

Categories
Books & Literature Technology

How I stay informed…

A few months ago, a lovely chap called Kavi Guptta asked me how I stay informed. I wrote way too much and he had to edit it down aggressively. For anyone who is interested, here are my full, unedited responses

Describe your daily reading habit:

My reading habits have changed a lot over the years. I did postgrad work in the humanities, so I read a huge amount of academic literature in my mid-twenties — to the extent that it almost started to feel like I was bombarded by words everywhere I went. And yet I liked writing and reading — they feel like scaffolds for thinking about things—so I kept up that tradition of over-consumption when I abandoned academia and ran away to join the Internet.

This was sort of the mid-nineties, and I was still doing it by the earliest days of weblogs in 1999 to 2005-ish. I had a ritual of waking up around 8am, spending about an hour on the web reading and writing on plasticbag.orgbefore heading into work—which thankfully was a pretty short trip—and then keeping up with various discussions and news sites throughout the day. This escalated even more when RSS readers started to become popular. I’d be trying to keep up with several hundred sites — mostly interesting individuals talking about technology and the web—but I’d also be a member of several online communities where I could stretch my legs in various other areas.

Probably three things changed my reading practices. Firstly, my work got much more involving and creative. By the time I was running a little R&D team at the BBC I found myself attacking increasingly meaty and interesting problems on a daily basis, and that pushed me into a very different mode. General reading started to fade away and started to be replaced with much more directed exploration around the things that we were building. That’s stayed with me. A lot of the time I stumble upon really odd new areas simply because one of the projects I’m working on involves me knowing about it. And because I really like the transition point where new technology starts to pushes into the mainstream, I’ve followed all kinds of areas from social software, media distribution, the web of data, location services through to today the Internet of Things.

Secondly, I started to realize that the conversations moving around the internet on a daily basis were starting to repeat themselves. My first exposure to the ‘are bloggers journalists’ discussion is now twelve years old — but it would reappear every three or four years as a topic of discussion again. For the people involved in them, they’re no doubt interesting and new but I started to realize that the larger trends in technology don’t actually change on a daily basis and that once you’re up to speed with the area you’re currently focused on, you barely need to keep half an eye on the daily chatter.

Finally, the ways that I could engage in conversations with my friends or follow complete strangers changed a lot. I pick up a lot from just being in a room with interesting people and letting that inspire me to wander off in strange directions. Today, things like Twitter mean that I can pretty easily absorb the general zeitgeist by dipping in and out of what interesting people are discovering or pointing at in the background of my day.

Now, a normal day for me probably starts with checking Twitter to see what my friends are up to, and then I’ll probably have it on in the background pretty much all day. I dip in and out when I’m bored or when something interesting appears to be happening and I probably stumble on most of my news that way. On my way to work I listen to podcasts from the UK like In Our Time. I glance at The Guardian every day, read MacRumourscompulsively because I find Apple fascinating, and spend a lot of time on iO9 to satisfy my nerd needs.

Beyond that I’m now subscribed to a few RSS feeds again, but they’re very different from ten years ago — they’re either weird fragments of the future, like BLDG BLOG or Future Drama or they’re deep design blogs where I can get exposed to beautiful things and extreme craftsmanship of various kinds. As I get older, I get more and more interested in the complexities and depths of craft — ink traps in typography, printing methods, modernist design and stuff like that. I love that dedication to detail and the way that your appreciation of something deepens when you understand why it is the way that it is. Otherwise I mostly focus on reading and research for particular weird product problems I’m having when I’m trying to assemble things in my head, although I do read a fair number of non-fiction books on my iPad and a lot of shiny, ostentatious, brightly-coloured comic books full of big weird ideas.

How do you learn?

The short answer is I learn too slowly. The longer answer is that it’s mostly pretty directed — I like collecting new skills, but I mostly collect them because I have something I want to do with them. As I said, I started off planning to be a humanities academic before I discovered the web and when I had discovered it—like many people of my generation—I sort of had to do everything. I had to get my head around the code side, the design side and what would now be called the ‘product’ side. I spent a lot of time writing for the web, designing sites, building them, and while I’ve had to specialize a lot over the years, I try really hard to be able to talk intelligently to designers, engineers, business people and product people as well as critics and thinkers outside my specific area.

At this moment in history it feels to me like a lot of things that were separate and silo’d have been put in the same territory (the internet) sometimes for the first time ever. And so it feels like the greatest creative potential can come from trying to see how those things or ideas can connect to each other to make something new. That’s sort of how I understand things right now — that the more things you know, the more connections between them you can generate and the more amazing stuff happens.

By way of an example, a while ago when I was first getting into thinking about the Internet of Things I had to recognize that I didn’t really know how to manufacture physical products. My business partner is good with electronics and soldering irons and stuff like that, but that seemed like a mountain of work for me to get into at this stage. So I followed the lead of a few of my friends and started exploring how you could mock-up videos of objects apparently in people’s homes using 3D software and After Effects. To give myself a goal, I decided I’d try and do three or four very short videos that I could use in a talk I was going to give. It was enormously painful and the videos are not very good, but they communicated the idea.

But interestingly, we also started working on a different project around that time focused on interactive graphics for TV. My week of fiddling with After Effects was enormously helpful in thinking about what the possibilities might be for a live 3D graphics compositing + a TV programme and we explored all kinds of fun ways of integrating live data into TV shows using the right color palettes, vanishing points, while being sensitive to edits and all kinds of other stuff. That little bit of exploration in one area just triggered all kinds of fun ideas (plus we got to work with Greg Borensteinwho wrote the book on hacking the Kinect, which helped). Short directed forays with specific goals into weird areas end up being really really useful.

Where do you go for inspiration outside
of your day-to-day expertise?

One thing that has been very clear to me throughout my years in tech has been how useful my abortive PhD was. It had absolutely nothing to do with tech in any way whatsoever — I was writing about identity politics, film theory and Ancient Greek tragedy. But it gave me a whole range of mental tools and ways of looking at the world — new ways of understanding that things haven’t always been the way they are today. I genuinely believe that you pick up weird mental or creative strategies throughout your life and that a widely read person who is keen to employ some of what they learn will always do more interesting (if not always more valuable) work than someone who is super deep in only one or two areas.

So I guess if I had to point to non-tech specific areas that give me some perspective or inspiration, they’d probably include history or literature or graphic design a lot, as well as fantastic TV shows, books and comics. I mentioned In Our Time earlier, but I can’t recommend it enough — forty five minutes each week of a completely different subject, from the arrival of trains into Britain, or Ancient Alexandrian Librarians all the way through to Dark Matter. Matt Webb introduced me to it years and years ago. I suppose the truth is that you never really know what might trigger an idea for you, so exposing yourself to a lot of different things is probably the best strategy.

One particular thing that I do find very good for inspiration though is Pinterest. I know that for many people Pinterest is all about collecting hairstyles or recipes, but there are millions of people on it, each of whom cares about something different. I follow people interested in typography, visual design, illustration, information visualization, maps, network-connected devices and many other things. It’s such a great way to trigger bits of your brain into doing something useful but unexpected. I use it every day…

What’s one piece of advice for someone who wants to do what you do?

I’m not entirely sure I know precisely what it is that I do, but I think more generally my advice would be to play a lot, explore lots of things, sketch lots of ideas, learn weird bits of software that interest you and do something with them — to fundamentally try and work out what it is that you actually enjoy doing and then to do more of it. For me it gradually became clear to me that what I liked more than anything else was making things that I could show to other people or that other people could play with or use. The internet taught me how much fun that could be, and since then that element of my daily work has stuck with me as the fundamental goal — at times I’ve gone astray from it (taking the money instead of doing work I can be proud of) and all that did is make me hate my job and start to hate my discipline. But when I get back to making something that I actually care about with people that I like, there’s genuinely nothing better, occasionally harder but definitely more rewarding.

Other than that, I’d say that if you can find a passion that’s fulfills you that actually pays the bills then you’re incredibly lucky. Not everyone gets to find that in their lives. Don’t hold back, jump right in and do it as much as you can.

If you’d like to see what effect a good editor can have on my writing, the original article is here: http://www.forbes.com/sites/kaviguppta/2014/04/25/how-im-informed-with-tom-coates-of-product-club/

Categories
Design Internet of Things Technology

In praise of boring objects

This piece was originally written for the Things conference in Berlin in May 2014. It was originally posted on Medium.

Unlike the moral universe, the long arc of technology curves not towards justice, but boredom. And—at one level at least—it’s boredom that we aspire to.

We might initially thrill at a world of new functionality, new abilities and new super-powers, but we also want and expect that initial strangeness and magic to be gradually digested down into the substrate of our lives — becoming a part of the infrastructure and background of our world, barely visible, unacknowledged, almost natural.

It only takes a moment to see how uncomfortable we are with the insane, sustained complexity of the world we inhabit. Simply look around and attempt to imagine the origins and histories—the family trees—of all the products, services and objects that you can see. When you do, the spiralling complexity of it all is enough to give you vertigo.

Each thing you glance towards started with the mining of minerals in far-off lands, the harvesting of crops, the breeding and culling of animals, the processing of materials, complex machinery that grids and boils and melts and refashions. In parallel, each object has an intellectual history – reaching back across the creative labour of two thousand years worth of thinkers whose work brought us to steam, electricity, resistors, computers. Material history and scientific progress are brought together in designs on computers or sketch pads, with human hands and minds stretched across the planet, co-ordinating a collision of components from which emerge functionality, form and colour. Still more diverse people across the planet are involved in the polishing, packaging and distribution of these things, before the great ships cross the seas or trace lines across the sky, carrying these objects and materials to new homes. Finally, these objects are arranged in all their staggering overlapping complexity on the street in front of you, or the office in which you work, or the home that you love — by thousands of interconnected yet independent individuals — each object bringing new abilities into the world and into your life, making things possible that were never possible before.

So why don’t we choose to live in this world of magic and wonder all the time? Why don’t we choose to be continually surprised and stimulated by the intricate patterns formed in every direction, stretching off into the distance? Why instead do we let that complexity disappear, dissolve into our daily experience? The answer is simple — because it would be overwhelming. There’s only so much space for magic and weirdness in our lives and we reserve that space for the immediately strange, or dangerous or novel.

The small amount of drama that we’re able to deal with each day is just a sliver, shiny brightly on the water. The great invisible iceberg under the ocean is the realm of the useful, of the everyday, of the objects that don’t continually push themselves front of mind, and that’s where products and services must eventually aim if they are to be fully integrated into our lives.


It’s an exciting time to be involved in the Internet of Things. It’s been almost thirty years since people started hypothesizing about a world of ubiquitous and pervasive computation, and twenty since the first toasters and coffee pots were connected to the internet. Over that time we’ve seen endless playful experiments and explorations of how objects might connect to each other and to us in thrilling new combinations of object + computation + network. But almost none of this work has reached the general public.

Only now is that starting to change, and fundamentally that’s because evolving technologies and interaction patterns have reached a sufficient level of maturity to finally let the potential unfold. You can see it everywhere — in the pervasive internet access that now spreads across the airwaves, in cheap networking components, in tiny computers stamped out by the millions and available for pennies, and in new interfaces in everyone’s pockets through which devices and appliances can be better controlled and understood. The world of the near future now looks like a place where every object can be connected to the network cheaply and easily. Finally it seems like the Internet of Things is ready to go mainstream.

More interestingly, this new generation of Internet of Things ideas seems qualitatively different from those that came before – they’re more in tune with the deeper affordances of the Internet. One very significant move has been the layering of the concept of multiple user accounts onto physical items — a now banal element of most web services is transformative when layered onto physical objects or spaces. Zipcar alone shows how some of the affordances of web services have the power to change how we interact with physical objects completely. A car is no longer an object that is owned, but a service that you can commission or spin up as needed.

And instead of adding great chunks of immediately out-of-date hardware into appliances with decade-long replacement cycles, much more of the intelligence is being abstracted out onto the network. To use a biological analogy, the objects themselves increasingly contain only eyeballs (sensors) and muscles (actuators). The nervous system and the brain are now elsewhere, abstracted out into the cloud, where they can be built upon and enhanced over the lifetime of a product. This service layer is becoming ever more as important as the physical object itself.

But for me the most important change is the move from IoT concept cars and interaction design experiments to a new world where the things we’re building are simply, cleanly useful. This latest generation of objects no longer ‘perform’ the internet—they’re not laden with touchscreens or web browsers or e-mail clients. They are instead just using the network to make better light switches, fridges and dishwashers — things that just make people’s lives a bit better.

Nest’s smoke alarms and thermostats are among the first of this new breed of products that are not simply concerned with showing off and instead are fundamentally concerned with making better objects using the network as a material. And other manufacturers are starting to follow in their wake, with large companies that produce every kind of appliance in the home, office or wider world starting to wake up to the possibilities. This is new, and as far as I’m concerned it constitutes the largest and most significant shift that we’ve seen in IoT for years.

These companies are looking for simple life-enhancing benefits from connected objects — whether it’s the tumble drier that turns on when you leave the house, the lights that turn down in any room you’re not occupying, or the fridge that informs you if it’s malfunctioning and schedules a service automatically. They’re looking at the clear value in the more effective management of energy across smart grids. They’re looking at ways to protect users from theft by making objects that simply stop working when they are stolen. And they’re looking at the new business models that become possible when you find new ways to share objects — whether it be the local home improvement tools that can be used by people only when they need them, through to appliances that you don’t buy but pay for by the use.

It’s everyday things like this that will get the Internet of Things out into the world. But it won’t happen by itself. Large appliance companies are no more able to turn their appliances into services than web technologists and service designers are able to assemble fridge-freezers out of JSON and iPhone apps. As never before, manufacturers and online service designers need to collaborate to build something collectively that is genuinely new — simple, useful, networked products that bring genuine value to people’s lives. And we need to help that collaboration happen, reaching out to each other as we do so, if we’re really going to make a world of useful, networked smart objects.


And that is the goal — a planet of objects writing and responding to the network, knitting themselves together in a new pervasive infrastructure that opens up endless new possibilities for all of us. That may sound like a leap for a discipline currently focused on super-evolved toasters and conversational light-switches, but think for a moment about where we go next.

The stage beyond this one is full of latent burgeoning potential that we cannot tap—or potentially even understand—until the network is pervasively woven into the world. Just as the social network couldn’t happen until enough people were online to make it interesting or useful, so new categories of services will start to appear as more and more objects come online.

In the near future, the world we’re building together is fundamentally about ovens and dishwashers, parking meters and water pumps. But the services we build today for our cars and phones and robot vacuum cleaners are together forming the bedrock for a tomorrow that (from here) seems much less predictable and far more exotic. The smart thermostat and the car-sharing services you use today are test-beds for the next fifty years, and the ideas we build into them now will unfold in fascinating ways that could have huge implications on everything from public space to the very concepts of ownership and sharing. Fundamentally, in our work today we’re starting the process of blending the physical territory of the world and the digital map of the world into something that becomes better than either could be by itself. Perhaps even one where the distinctions between the two no longer seem relevant or useful.

By the time we get there, it may all seem very obvious, prosaic and — grudgingly we can admit — probably profoundly useful. But for the moment this sliver of light on the water is unknown, untapped, unprocessed and fascinatingly weird and new. And we have time to appreciate that wonder, to taste it, to bring form and shape to it, to dream and argue about it, even as we day by day build towards it. The future, at its best, is a site of endless potential and change. And while in the end, technology might tend towards boredom, the creative joy is getting there.

This piece was originally written for a free publiction given out at the Things Conference hosted in Berlin in May 2014.

Categories
Politics Technology

On being adult about childish behaviour…

I promised myself I wouldn’t get involved in the whole debate about Adria Richards, but as it drags on and more and more articles are written about it, I find myself troubled by the extent of the polarisation going on.

On one side we have the people who are arguing fundamentally that Adria Richards over-reacted to what appear to be a number of clumsy and childish jokes – jokes that (from what I’ve read at least) appear to be of the ‘he’s got a big dongle’ variety. My opinion? Yeah, I think she massively over-reacted, made a performance out of the whole thing that was entirely unnecessary and got a couple of people told off at a conference who maybe over-stepped the line but probably didn’t deserve that kind of public kicking.

On the other side we have the people arguing that Adria was treated absolutely abhorrently afterwards and that the sheer depth and violence of the things said about her and thrown at her – the volume of the vitriol she had to experience – went way beyond blunt sexism. These people are arguing that this demonstrates structural misogyny, clear discrimination, the horrible consequences that meet a woman who prominently stands up and makes a stand. My opinion? Yeah, that all makes sense too. She was indeed treated awfully. All of us working in the industry should be ashamed of the whole situation. It’s been hideous, horrific. Beyond that. It’s morally wrong.

But here’s the bit that puzzles me. Both of these positions seem to me to be entirely correct and both of them seem to be completely compatible as well. It seems to me to be a vanishingly small proportion of people under fifty who would be legitimately offended by a big dongle joke. It also seems to me to be a tiny proportion of the people in the industry who seriously think that death threats, rape threats and massive sexist comments are something to be encouraged.

So why has it all become such an ungodly fight? We seem to have approached a point where any actual sensible discussion of questions raised by this situation is borderline impossible. The positions are polarising to such an extent that—rather than just accepting what the vast majority of us must surely know to be true—everyone’s being pushed, or pushing themselves, to the edges. The arguments now appear to be that either Adria fucked up and for this reason she deserved to get rape threats, or that since she got rape threats she cannot possibly have fucked up.

These are both ridiculous positions! These are insane positions! These are totally irrational positions! In our attempts to find meaning in this event we’ve got people trying to find a neat narrative that wraps everything up elegantly and cleanly. But such an attempt is doomed to fail here. No one comes out of this cheerfully. There is clumsy human self-importance on the one side and a great swathe of unpleasant, unwashed, dickish, abusive and disgusting morons on the other.

Now don’t get me wrong. I’d never for one minute suggest that Adria’s actions and the responses to them are commensurately awful. The horrific attacks on her remain beyond the pale, where her behaviour (in my opinion at least) was merely a bit eye-roll worthy. It’s just important to remember that the extent of the venom she’s had to endure does not in itself make her an infallible saint – any more than the error she made in any way justified the demonisation she’s experienced.

The world is more serious and flawed and nuanced than these narratives would make us believe. The tech industry is not 90% full of sexist arseholes desperate to drag all women down. Nor is it a magical meritocracy in which all the right people achieve riches. Over-simplifying things to this degree makes it harder to solve the problems that we have as a culture. It makes it harder for us to fix things. We can’t afford to just react like this. It helps no one. We have to be adult about childish behaviour.

Categories
Business Politics Technology

Should we encourage self-promotion and lies?

A couple of days ago, Clay Shirky wrote a piece on his blog called A Rant About Women which took as its subject the comparative comfort with which some men are prepared to market themselves, mislead and lie to get ahead compared to women.

I’ve been reading responses to this piece on Twitter and elsewhere, and I’ve become increasingly horrified by what I’ve seen. Generally, it’s being viewed as a call to arms to create a new breed of women who are as self-important, self-promoting, shameless and arrogant as some of the worst (and most celebrated) men in the industry. This attitude is being viewed as the ‘way to get ahead’ for any individual wanting to make their mark in the world.

I’m prepared to accept that there’s a correlation between attitudes to competition and self-promotion and gender. I’m not as prepared to take it as far as Clay seems to, but I’ll go along with its generalised existence.

And clearly, if aggressive self-promotion and pompous self-aggrandizement is what gets people ahead in the world, then at the individual level, it’s better to perform in that kind of way than it is to sit passively and watch yourself get passed over by more clumsy, venal, smug, aggressive, macho idiots.

But at the level of the company, at the level of the community, at the level of the industry – are these attributes in fact in any way desirable? Does self-promotion really lead to great products or projects? Is the ability to lie and mislead really what it takes to achieve?

My experience has been that there’s definitely a role for the arrogant and the pushy in the creation and promotion of a project. It’s also taught me that this skill is a small part of the set of skills necessary to produce something great.

The kinds of things that result in great products are tangible skills, a desire and a pleasure in collaborative building, an aspiration and sense that you’re making something important, a sense of teamwork, room to experiment, the ability to bring out the best in the people around you, a good work ethic.

Alongside that a desire to show-off can be really beneficial, a confidence in your ability is essential, the ability to push yourself into new areas certainly a benefit. But these attributes can also get in the way. There’s something in American culture in particular which values the pushy and the determined, but we’ve all worked with people whose confidence massively outstrips their abilities, who cannot work together with other people because they think they’re superior to everyone else.

And we’ve also met a whole bunch of people in the industry who do nothing but self-promote, working day and night to sell themselves, and achieve positions massively disproportionate to their tangible abilities. There are people in our industry in positions of substantial power whose reputation is built upon the way in which they present themselves as being visionaries and experts. Some of them have found that it’s simply more efficient for them to spend their days building that reputation through PR and self-promotion than it is to demonstrate it through the things that they make, the value that they create.

I’d never argue that we should forcefully reject anyone who manifests confidence, skills in self-promotion or who is cocky enough to sell themselves. But what I want to strongly resist is the idea that it is these attributes that we should be promoting – either in women or in men.

It should be unacceptable for us to say that lying about one’s abilities is something that everyone has to do to get ahead. It should be unacceptable for us to say that arrogance and aggression are to be aspired to.

Instead we should be demonstrating that great projects, like the ones Apple produces, are at least in part based upon trying to produce the best thing possible, feeling the integrity in the product you’re making. Trying to do something good. We should acknowledge the example of Flickr who created an astonishing culture of extremely talented engineers and designers around the very real aspiration to make something beautiful, powerful and good for the world. Or the guys at Twitter who discovered their idea initially by letting small groups experiment in interesting directions rather than dogmatically following the vision of a bold cocksure individual.

Good projects come from good people, good vision, good execution, good collaboration, good insight. And it’s these traits – and the ability to spot them – that we should be encouraging in our colleagues.

The right thing to do is to get it into the heads of our VCs and companies that a hunger to win at any cost is not the main attribute of a creative or productive person. That the ability to be intelligent, think through problems, work with other people, develop ideas effectively – that all of these traits are better indicators of success than how big they tell you their testicles are! That the person who comes to you with the biggest pitch is not necessarily the person you should be listening to.

And while encouraging people to spot the talented and the creative, we should also be considering how we shame those people who self-promote without creating. The financial collapse has taught us that rhetorical bubbles divorced from reality are a danger to us all. We’re already approaching this point – our industry has become venal, insular and dominated by marketing. We have come to value the wrong things. And if we want a continued vigorous, creative, free, open and equal environment, that’s something we have to fix. It’s not something to aspire to.

Categories
Journalism Politics Science Technology

On the OLPC Movement…

A couple of months ago I was asked by Icon Magazine to write a review of the OLPC XO laptop for the developing world. You can read the finished article in their January issue or on their site (OLPC review on Iconeye.com). However, since then, I’ve been thinking a lot about the context and background to the review I wrote and decided that I wanted to elaborate around it a bit. I’ve also wanted to put up a version in public that hadn’t been edited for length (however judiciously). Hence this post – firstly some background to the piece and then the piece itself.

When I read a review of the XO I expect to stumble across some fairly standard positions. I expect the article to question whether the developing world really needs a laptop. I expect them to talk about the ecological impact of these laptops. I expect them to decry the project as (at best) utopianist folly and (at worst) some form of western naïve semi-colonial oppression. Most of these arguments make no sense to me at all.

The first position seems to be based on the assumption that the people of the developing world be better off gradually developing their economies through farming to manufacturing and ending up gradually in high technology industries. I personally think this ideology dooms these countries to always playing catch-up to the west. If there’s any chance of them leap-frogging great swathes of industrialisation to create a working and creative population that can compete on a world stage, then I don’t understand how any of us could stand up in good conscience and decry it.

Environmental damage through a proliferation of laptops seems to me to be probably indisputable, but what’s the alternative? Is it fair for rich countries to consume vast amounts of resources but stop poor countries having access to the same services ‘for the good of the world’? Can we really in good conscience deny other people what we take for granted? Perhaps if we were sending out hundreds of millions of Macs or PCs there might be an argument here, but the XOs are massively less damaging to the world than any of those devices.

The utopian accusation may have some truth to it. It’s difficult to know precisely how much chance a project like this has of success. And it’s difficult to know whether the technologists behind it are busily projecting their own ideologies onto developing countries in defiance of what those countries actually want. Probably the only way we’ll find out for sure would be to provide the machines to a few disparate groups of young people across the world and see how they develop–see what opportunities it opens up for them. Personally, I find the arguments convincing. I think there is a net benefit to come out of this. I think it will help. But it’s pretty tricky to distinguish your own beliefs from your prejudices. I wouldn’t blame anyone for not being so confident.

For me, it comes down to the way we want to operate in the world. It’s extremely easy to adopt a pose of scepticism and cynicism about any attempt to change things or push them forwards. I’ve said before about a particularly aggravating tech commentator that naysaying is a sure-fire way to look sensible and intelligent without any of the effort of actually having to think. I stand by that, and I think the OLPC project has had its fair share of this kind of thinking.

Personally though, I believe that it’s possible to work for the good of all and improve the world. I think it’s a decent and honourable thing to apply whatever means you have at your disposal to raising the aspirations and possibilities of one of the planet’s most squandered resources–its residents. And I do buy the geek rhetoric that access to information, communication and education cannot but help people. As such, I’m prepared to give this project and others like it, the benefit of the doubt. And that’s why I decided to write this article in this particular style. I hope you enjoy it:

There’s something grotesque about reviewing Nicholas Negroponte’s XO–the so-called “$100 dollar laptop”–for a magazine like Icon. And that I’m writing the piece on my gas-guzzling SUV of a MacBook Pro can only compound the horror. This is not a machine designed to be evaluated by people like me. Nor is it meant to be bought by the kinds of people that will read this magazine. To talk about it in the same design terms as a lamp or a set of headphones borders on criminal, because in every way that really counts the XO is not a consumer artefact. It’s not trying to wheedle itself into your living room. In fact, quite the opposite. It has more in common with a clean water pump or an honourable approach to third-world debt than it does with an iPod. It’s a sincere but radical political act.

The result of a two-year project by “One Laptop Per Child” (OLPC), the XO aims to introduce primary school children in the developing world to the educational possibilities of technology and the network. Green and white with a tough, textured plastic body about the same size as a lunch-box, it has been optimised in every way to deal with the extreme conditions of its use. Its astonishingly frugal use of electricity allows it to function in areas where power is sparse or even non-existent. The screen switches into an energy-efficient black and white mode that is also readable in direct–even aggressive–sunlight. The rubberised keyboard seals the device against dust and water. Even the friendly green “ears” of the device serve a triple function – acting as latches, protective shields for USB ports and as antennae designed to extend the range of the distributed wifi networks that will connect children across the planet.

And this is a device optimised for the young. The keyboard immediately reveals the clumsiness and size of fully-grown fingers. Each key is springy and responsive–fun to touch and explore–but they’re packed tightly together to help small hands roam effectively. In every dimension, the XO is child-shaped. The grasp of the handle, the heft of it in your hands, the way it swings when you walk–it’s enough to make any adult feel like a freakishly large mutant. And it’s not only child-shaped, it’s child-resistant – it feels resilient, solid, indestructible–as if it could be used as a tennis racket without sustaining any real damage.

Yet what’s truly extraordinary about the XO isn’t the way it’s been tailored to work under extreme conditions, but the bets it places on our collective political and creative future. Geek utopianists have infused every aspect of the device with their own profoundly aspirational, positive and humanist political ideology. The XO is their lever to effect change at a global scale.

You can spot it everywhere. Every aspect of the device — from the operating system to the mesh networking that distributes connectivity to each machine — works on the principle that each node on the network can accomplish more together than they can apart. Every application on every machine is designed to operate in a social context – you can show off your work, share your web browsing or advertise an ongoing discussion. Some applications–including a version of Connect 4–are only functional at all if you have other people to play with.

The collaborative, communal experience is tied together by the “zoom interface” – the XO’s version of the Finder or File Manager. It allows a user at any time to zoom back from one particular application to their desktop, then to their community of friends and then still further to see everyone on the network. While zoomed out, you can see clumps and clusters of people collaborating and playing, always connected and situated within their community. The XO is not a device for loners. It is a device that believes aggressively in society and aims to support it.

There are also challenges to our traditional understanding of intellectual property. The communities in the developing world that cannot afford life-saving drugs can find themselves similarly constrained by the cost of textbooks–and often for similar reasons. But with a turn of the screen, the XO becomes a simple ebook reader connected to a network. It’s an environment ideal for the distribution of free knowledge, and so it’s no surprise to see Wikipedia involved in the project. Information, as the technologist’s mantra goes, wants to be free–and the XO is there to help that happen.

You can see similar principles at work in the pervasive use of open-source applications and software like the Firefox browser or Linux. This software is free to use, install and distribute but–more importantly–offers its very code up to exploration and change. The XO revels in this opportunity, making it easy for children to access and edit the very software of their machine. There are no finished creative works here, but simply sites for continual exploration and learning.

In every area, this iconic object is an attempt to refashion the world in the image of the dreams of its creators – noble, vigourous, creative and expressively utopian dreams. Every element is impregnated with these aspirations of sociality, play, creation, freedom. As a project and as a device, it’s beautiful and revolutionary.

If the perfected whole succeeds in its mission, these aspirations may find a new home in the minds of generations of children in the developing world. And this new generation – growing up able to access and manipulate knowledge, technology, literature, music and code – will bring to the networked world their new perspectives, voices and needs. It’s a project to transform the world: this small device has a substantial mission. It’s not a laptop, it’s a movement. And it deserves our full support.