Adam Curtis is a filmmaker who intrigues and frustrates. His Century of the Self and Power of Nightmares peeled back the layers on Freud and modern capitalism, and the rise of neoconservatism and fundamentalist Islam, respectively, in a new and interesting light. Curtis may not be right, he may not even be telling the whole story, but he offers an angle, a way of skewering and unruffling our preconceptions. However, with his 2007 The Trap, he started falling off his usual run of form. It offered a frustrating take on the modern take on liberty, from positive to negative, and the perniciousness of game theory, behavioral economic models and performance targets – connections a little too technical and forensic to explain just with mashup of videos and laconic voiceover. A waffly final third culminated in a call to arms for positive liberty, at odds with his usual dispassionate tone of voice.

Four years on, we have his new series, ALL WATCHED OVER BY MACHINES OF LOVING GRACE, and the chance to see whether The Trap was an aberration from true form. Like its predecessor, Curtis delves into the technical not just the historical. His basic thesis (I’m summing up) is as follows: the selfish Objectivist philosophy of Ayn Rand inspired a generation of Silicon Valley geeks to create computer systems with the aim of removing the shackles of government to create a utopia of free individuals; these same system are then used by the creators of the New Economy (led by another Rand acolyte, Alan Greenspan) to create a new economic miracle controlled by the banks, Goldman Sachs and the Federal Reserve. But instead of creating a utopia they created chaos – the machines they had such faith in failed, and despite producing an economic crisis in Asia in 1997, we kept faith in them to create an even larger worldwide crisis a decade later, from which we can see no way out of.

The problem with this is that it’s seeing too many links where there aren’t any. Not every Silicon Valley company was inspired by Rand (Curtis named one or two examples at best, neither of who leading lights on the scene), and the area’s philosophy owes arguably more to the countercultural movement and political climate of California than Rand’s self-indulgent miserabilism. It’s certainly a long way away from the conservative, East Coast, market fundamentalist philosophy of Greenspan, Robert Rubin and indeed the entire neoconservative/Chicago School generation of politicians, economists and bankers who eventually assumed political control. Cyberlibertarianism envisions a future of individuals networked together, free of hierarchy or even the state; market fundamentalism celebrates harnessing the aggregate of individuals’ behaviour for greater prosperity and stability. In short, one coast’s philosophy created John Perry Barlow, the other Alan Greenspan.

Curtis’s other flaw is to confuse “machines” with what machines actually run. A computer is just a unit for processing numbers in any number of ways. They are just boxes, glorified calculators. It’s the software we run on them that makes them do “evil” and this software is made by human beings. Having spent so long talking about a generation brainwashed by Rand, Curtis now attributes all the evils to the machines. But who programmed them? Who first thought of using them for automated trading, just-in-time manufacturing, supply chain management and all the things that are now taken for granted in the New Economy? For a storyteller who loves to peel apart the unknown and the people behind history, Curtis instead frustratingly wastes his time on peripheral figures such as Ayn Rand (who died a recluse in 1982) and Bill Clinton (distracted by the Lewinsky affair and powerless to stop the SE Asia crisis), rather than the people who built and shaped the information economy.

The result is a mess, with Curtis making oversimplified and hurried connections between various subplots. But that doesn’t mean there isn’t a story to be told amongst this clutch of different tales. How did a bunch of so-called geeks and slackers, growing from the midst of the counterculture, create a multibillion dollar paragon of capitalism? How did the conservative, sleepy institutions of Wall Street become seduced by the wonders of technology and grow hypertrophically on computer models, automated trading and complex financial instruments? In short, how did Barlow and Greenspan’s generations become allies, intertwined and taking on each other’s aspects and practices? And finally, how have we become so dependent on these systems, making them become so ubiquitous and invisible, that we didn’t notice things were going badly wrong until it was too late?

If you think this sounds familiar, it’s because Curtis used this intertwined-dichotomy style of filmmaking so well with The Power of Nightmares: the story of how the ideological descendants of Leo Strauss and Sayyid Qutb ended up as putative enemies, yet neither could live without the other, and both were grounded in the same similar grievances with individualism and liberalism. It’s odd that Curtis was able to portray the balance of similarities and differences in that film, yet with AWOBMOLG he struggles to make sense of it all, and ends up merely telling the “what” rather than the “how”, giving us numerous red herrings in the process.

It’s easy (but patronising) to say that’s because “technology is hard” and it’s difficult to comprehend it and history together rather than history alone. But Curtis is not a stupid man. It’s perhaps more charitable to say it’s easy that when it comes to relatively-uncharted history of the information age, there is simple so much more information and so many possible narratives, it’s easier to see pattens where there are none than not. But, this was just part one, and maybe parts two and three are better, and a lot more coherent.

In the meantime let this not detract from Curtis’s earlier works – if you haven’t seen them, The Power of Nightmares and Century of the Self are both available from His KABUL: CITY NUMBER ONE blog is a collection of blogging and archive clips about the Afghan capital, and well worth reading. And I retain a soft spot for It Felt Like A Kiss, an avant-garde experimental attempt at storytelling based on 20th century history commissioned by the Manchester International Festival. AWOBMOLG was disappointing but don’t let it put you off entirely.

The curious case of Twitpic’s disappearing Terms of Service

Update: Thanks for all the RTs everyone. For those of you who don’t want to read the whole story, the TLDR version is this: Twitpic changed their ToS to restrict users from selling their uploads to agencies, then retreated very hastily after a Twitter backlash. If you want to know more, read on.

Another week goes by, another scrape on Twitter. This time slightly less interesting than my stumbling on the bin Laden liveblogger, but still intriguing, as I have never before seen a tech company edit its Terms of Service live in front of my eyes, like a Wikipedia entry, in response to a Twitter storm.

Twitpic is one of the leading Twitter image hosting services out there – about four million daily users, the site the famous photo of the plane landing in the Hudson River first appeared on, and the default image service for the Twitter iPhone and desktop clients. It has become big business – it brings in $1.5m in ad revenue a year.

If you run an image hosting service, you have to be careful in how you treat users’ copyright. Your users (usually) own the copyright to the photos they upload, but the service will need some form of non-exclusive royalty-free licence to legally host it on its servers. This licence is included deep in the terms of service. So far, so dull. However on May 4th, Twitpic’s terms of service changed, specifically the copyright section. The copy I have for reference is from Google’s cache and is mirrored here and it’s the first four paragraphs of the copyright section that bear most interest:

By uploading content to Twitpic you give Twitpic permission to use or distribute your content on or affiliated sites.

You may not grant permission to photographic agencies, photographic libraries, media organizations, news organizations, entertainment organizations, media libraries, or media agencies to retrieve from Twitpic for distribution, license, or any other use, content you have uploaded to Twitpic.

All content uploaded to Twitpic is copyright the respective owners. If you publish content uploaded to Twitpic on the web for personal and noncommercial purposes you are required to link back to the original content page on Twitpic and attribute credit to Twitpic as the source where you have taken the content. For example a Twitter “retweet” is acceptable provided the original content link on Twitpic is what is retweeted. It is not acceptable to copy or save another user’s content from Twitpic and upload to other sites for redistribution and dissemination.

To publish content for any commercial purpose or for distribution beyond the acceptable Twitter “retweet” which links back to the original content page on Twitpic, whether online, in print publication, television, or any other format, you are required to obtain permission from Twitpic in advance of said usage and attribute credit to Twitpic as the source where you have obtained the content. No user may grant a third party permission to copy or save content that has been uploaded to Twitpic.

The first part highlighted is a clause seemingly denying anyone who uploads a picture to Twitpic the media exploitation rights for that picture; it specifically targets those businesses who might want to pay for it. The second is a more vaguely-worded catch-all clause that, in the most draconian interpretation, could deny a user from uploading their own pictures to other hosting services like Flickr.

For comparison, this was the equivalent section from the terms of service in May 2010, which is the most recent copy held on – sadly I have no more recent copy to compare with:

By uploading your photos to Twitpic you give Twitpic permission to use or distribute your photos on or affiliated sites

All images uploaded are copyright © their respective owners

[Note: I am not a lawyer, and this is just a lay reading of the situation. But it will become clear, I hope, that these passages are at the centre of what went on in this kerfuffle]

Although these changes were made on the 4th, having done a little detective work on Google Realtime, it seems no-one picked up on them for six days. A single Tweet by @JMRooker on the 5th noted they had been updated, but not on what had changed. It wasn’t until this article (in German) appeared in in Der Spiegel‘s tech section that noted the change this afternoon. The first Tweet I have found was by Beate Clever in German at 15:54 UK time; at 16:49 it was Tweeted in English by Oliver Reichenstein.

Although it took six days for the news to get out, once it did it spread very quickly. Oliver’s Tweet was spotted by my We Are Social colleague Hannah, who retweeted it herself at 17:59, and I promptly followed at 18:05. By this point the story had legs – though I was by no means the only person talking about it, my Tweets on the subject got picked up and retweeted by some excellent & influential Twitter people such as Zoe Margolis and Tom Coates.

You could be easily convinced this was just another angry Twitter mob. But those who responded shared some interesting points of view. Was I being unfair on Twitpic? Did the terms only apply to the version they hosted? But if so, what makes that version different from the original copy you took? Did it only proscribe you from sending the Twitpic to a media agency, and if so would emailing a separate copy of the image as an attachment be just fine? Or was it just a protection from unfair infringement by media organisations? With the new terms of service as vaguely worded as they were, it was subject to various interpretations as the buzz spread.

The specific mention of picture agencies coincided with yesterday’s news that Twitpic had signed a deal with the photo agency WENN to represent celebrity pictures posted on the service (what Der Spiegel picked up on). So this was perhaps a foray for Twitpic to become a citizen journalist version of the PA, providing free hosting in exchange for the right to licence the rights to picture agencies. The idea has been mooted before. In theory, the next time a plane lands in the Hudson River, Twitpic’s ToS would allow it to sell the rights to witness photos uploaded to the service, and prevent the photographer from seeking those rights herself, if the company so chose. Whether Twitpic seriously thought of this as a future business model, or was just enabling these terms now in case they would come in handy in the future, we don’t know, and I am not saying one way or the other.

Finally, apart from legalities, is what they’re laying out morally fair? The new ToS brought a fair bit of opprobium (not least from myself). But, do remember, with free image hosting, you aren’t paying, and although they are getting cheaper, bandwidth and scaling up a service do cost. Online display advertising is just one business model and has increasingly tight margins, so Twitpic may be in the early stages of exploring alternatives. As with any service offered online for free – always caveat emptor.

Twitpic managed to backtrack very quickly. At 18:45, the support team hurriedly tweeted back to me (and then others in the story, such as Oliver) the same message, stating:

@qwghlm We’re working on a clearer version of our ToS now to show better that we are not taking your copyrights or selling your photos.less than a minute ago via web Favorite Retweet Reply

They were true to their word; suddenly whole chunks of the text started to disappear from the Terms of Service page as it was being edited; I managed to take notice and livetweeted the progress. Round about 19:05 UK time, out went the second paragraph about photographic agencies entirely. At the start appeared an entirely new paragraph affirming copyright holders’ rights to their work. The paragraph on non-commercial reuse was cut, save for the final sentence (“It is not acceptable…”), which was merged with the new first paragraph. By 19:52 the final edit had been made: from the fourth paragraph, the third party sentence was cut out entirely, and the start was reworded to affirm it applied to reproducing other users’ content, and did not cover your own. In the updated Terms of Service, the equivalent text to the above now reads:

All content uploaded to Twitpic is copyright the respective owners. The owners retain full rights to distribute their own work without prior consent from Twitpic. It is not acceptable to copy or save another user’s content from Twitpic and upload to other sites for redistribution and dissemination.

By uploading content to Twitpic you give Twitpic permission to use or distribute your content on or affiliated sites.

To publish another Twitpic user’s content for any commercial purpose or for distribution beyond the acceptable Twitter “retweet” which links back to the original user’s content page on Twitpic, whether online, in print publication, television, or any other format, you are required to obtain permission from Twitpic in advance of said usage and attribute credit to Twitpic as the source where you have obtained the content.

Those changes might be a bit much to get your head around, so to make it easier to appreciate how big they are, I have created a side-by-side comparison of the two wordings – old is on the left, new on the right – and created a diff of them (the changed bits in blue). Feel free to click through to a full-res version.

Diff of change in Twitpic's terms of service

You can see for yourself that the changes are quite extensive, and the terms are now vastly stronger in their affirmation of uploaders’ rights. A mere clarification this is not, in particular the second sentence of the first paragraph – “The owners retain full rights to distribute their own work without prior consent from Twitpic” – is in marked contrast to the original wording.

Twitpic have published a blog post called “Your content, your copyrights“, which states their reasoning for changing the Terms of Service:

As we’ve grown, Twitpic has been a tool for the spread of breaking news and events. Since then we’ve seen this content being taken without permission and misused. We’ve partnered with organizations to help us combat this and to distribute newsworthy content in the appropriate manner. This has been done to protect your content from organizations who have in the past taken content without permission. As recently as last month, a Twitpic user uploaded newsworthy images of an incident on a plane, and many commercial entities took the image from Twitpic and used it without the user’s permission.

It’s great that a hosting service is explicitly protecting its users from exploitation and unlicenced copyright infringement. But the original draft of these updated terms of service made it clear that it was about more than just preventing unscrupulous news organisations from misusing photos. The now-removed sections were clearly about preventing users from selling their rights to their own uploads to third parties, not protection from theft. And unlike the apologetic blog post on May 10th, these initial changes to the ToS were not publicised to the wider community when they went up on May 4th; as I have detailed above, it took six days before anybody actually noticed – and by the way, all the credit should go to the tech team at Der Spiegel for spotting it. How Twitpic went about this change is not how a tech company should publicise changes to its users; whatever their motives were for updating their ToS or whatever plans they have for their business model, at the very least this was a major failure in communication.

Twitpic do deserve some plaudits for reacting quickly to the situation, answering those of us who questioned it on Twitter, and updating the Terms of Service to something more acceptable in ludicrously quick time and manner. I say more acceptable. They do still retain a licence to distribute your content as long as it is done in connection with their business, and that business model could well change from being an ad-supported image host in future. There’s nothing stopping you from exploiting the rights to your image, but they have those rights too. You may be fine with that, in which case carry on, or you may not feel entirely comfortable, in which case you may still want to choose somewhere else to host your images. I’ll repeat what I said above, when it comes to free stuff, caveat emptor, especially if you reckon you stand a chance of one day being the next person to snap a plane in the Hudson.

Update (18/05): Tom has received a letter confirming Twitpic’s intentions to “exclusively” resell photos through the WENN picture agency. There is still no announcement of this on Twitpic’s own blog or in the terms of service. So, what gives?

Tweeting the killing of bin Laden: how a little geekery and I (maybe) helped break a story

As the biggest news story of the week, the killing of Osama bin Laden, broke, I was on holiday in New York. As the clock ticked passed midnight local time (EDT) on Sunday night, my girlfriend Maha, sitting next to me on the sofa, passed me her Blackberry and showed me a retweet of curious remark related to the events unfolding. A Pakistani journalist, Mosharraf Zaidi, reminded his followers how he had earlier remarked:

:) RT @silicon_d: Mad props 2 @mosharrafzaidi for sixth sense: “What was a low-flying heli doing flying around Abottabad Cantt at 0130 hrs?”less than a minute ago via TweetDeck Favorite Retweet Reply

The television news we were watching had nothing new to show by now, and was resorting to reruns of President Obama’s address earlier. So, with my curiosity piqued, I started looking up to see if there had been any coverage of helicopters in Abbottabad earlier that day. Googling around normally found the odd news report about a possible training accident, but very little of substance or interest. So I turned to searching Twitter, specifically with Google Realtime, which allows you to exclude Tweets from before or after a certain time of day. This was important, as once President Obama had disclosed the location, Twitter exploded with mentions of it and it became impossible for ordinary Twitter search to cope.

With anything after 11pm Eastern Time excluded, I was able to find Tweets by a guy called Sohaib Athar, or @ReallyVirtual. Once I clicked through to his timeline, I found out he had actually liveblogged the entire raid, unaware that it was America seeking its public enemy number one. At 12.38am, I tweeted, and Maha tweeted too:

Wow. Turns out at least one person, @ReallyVirtual, inadvertently liveblogged the raid in Abbottabad earlier today than a minute ago via Nambu Favorite Retweet Reply

.@reallyvirtual appears to have liveblogged the raid w/o knowing it. go read.less than a minute ago via web Favorite Retweet Reply

And then at 12.41, three minutes later, Sohaib tweeted the defining moment of the story:

Uh oh, now I’m the guy who liveblogged the Osama raid without knowing it.less than a minute ago via TweetDeck Favorite Retweet Reply

(Tip: Twitter annoyingly displays Tweets’ times in a “x minutes/hours/days ago” format, but if ever you want to check the exact timing of a Tweet, hover your mouse over the that bit and a tooltip will give you the exact date & time, in your timezone)

By the next morning, Sohaib was one of the most famous Twitterers around, being interviewed on television and getting mentioned in most mainstream media outlets. His follower count shot up from 750 to just over 100,000 as of today.

Steve Myers of the Poynter Institute got interested in how the story spread and did some investigating (including talking to me and Maha), producing a phenomenal forensic blog post – and from his investigation it appears that mine & Maha’s Tweets were one of the first ones to mention him and may have broken the story.

Caveat: I say may – correlation does not imply causation. I looked on Google Realtime for earlier Tweets from anyone linking to his account pointing and highlighting his liveblogging, and could not find any, but that doesn’t mean there weren’t any. Nor am I trying to take too much credit for breaking the story – had I not tweeted about him, someone else would have found him sooner or later – the tools were there, and reasonably well known in the trade (and if you work in the media and don’t know them, then for God’s sake learn them).

Steve’s piece is a great bit of detective work and social network theory in one, and I’d like to pick up on a few points he made. Firstly, he states “the number of followers doesn’t matter as much as who those followers are” – this is a really interesting one and worth elaborating. I have about 3,300 followers on Twitter, but most of those are UK-based and would have been fast asleep. Maha has 1,300 followers, but she is a journalist based in the US, specialising in amongst other things, Pakistan. Her following, while smaller, is full of journalists, policy people and those with an interest in Pakistan and the wider region; these would be the exact kind of people to pick up on the significance straight away.

The numbers then don’t always count. But what definitely does count is the story. Steve picks up on the role I played bridging different social networks (in a paragraph that makes me feel odd, being referred to by my surname…):

Applegate was a bridge too, in a slightly different way. He added essential information that resonated with people and spurred them to pass it on.

I didn’t regard myself as a bridge at the time. I just thought, and tweeted: “Wow”. But then as it unfolded more it became clear that the unwitting Tweeting was a central factor in the story. Abbottabad is a relatively quiet town, populated by retired generals and known for its schools and universities (not to mention their military academy). Sohaib himself had moved there to get away from the much more dangerous and turbulent Lahore to find quieter climes – his Twitter bio states he is “an IT consultant taking a break from the rat-race by hiding in the mountains with his laptops.”

Given a popular narrative of Bin Laden hiding in caves and the like, to find out he was living in a mansion somewhere so quiet, so genteel and so near to the heart of the establishment came as a surprise. The key thing that made Sohaib’s liveblogging from earlier in the day so compelling was that it was completely unwitting, mirroring our own disbelief that Bin Laden had been quietly residing in the Pakistani equivalent of Tunbridge Wells all these years, without any of us knowing. The story chimed perfectly with our own emotions. And because the story had been unwitting, it was also candid and honest, cutting through the hype and speculation that the 24-hour news stations were resorting to.

Finally, the whole episode shows how transformative Twitter can be. As the story matured and his fame rose, Sohaib took on the role of citizen journalist, becoming a correspondent of sorts (not many other residents of Abbottabad are on Twitter, he remarked, it’s mostly Facebook). He conducted interviews on television, and ventured out into town to take photographs and report back on the mood in the town.

This is a far cry from the cynical caricature of Twitter as an echo chamber – a place where nothing new is said and everything is relentlessly retweeted. As the story progressed, Sohaib came to the wider community’s attention and it in turned shaped his role in the affair. His relationship with Twitter evolved – it went from being a place to remark on the events that had taken place, to realising their significance, to realising his own significance, and then finally embracing it with intrepidness, intelligence and good humour. I might have been one small factor that sparked the process off, but I definitely can’t take any credit for the phenomenon he has become – that’s entirely to his own credit, and something that we should celebrate.

Update (05/05): Maha has also blogged about the events, if you’re interested.