Synthetic media: be as afraid as you ever were

Synthetic media is a term I’d not come across before hearing about Google’s paper on fighting disinformation that was published this week. It describes those eerily realistic images and video generated through artificial intelligence (AI) or machine learning (ML) techniques.

You might have seen the video of the Jennifer Lawrence press conference where she has been gifted Steve Buscemi’s face as an example of this.

Image result for jennifer lawrence / steve buscemi

Deep fakes are what we more commonly call these.

Google says it is sharing datasets of synthetic media so that others can use it to spot deep fakes and develop systems that can do this. However it admits that there are limits to what tech can do – always slightly galling for a tech giant – and says it will need to also work with “researchers, policymakers, civil society, and journalists around the world”. Although none of these watchdogs have enjoyed complete success dealing with disinformation generated by non-silicon based actors, to coin a phrase. This last ditch defence against the tidal waver of fakes has been breached many times before.

Check your anti-tech moral panic

While we are trembling at the prospect of future synthetic media horrors, we might also recall that manipulating and misleading through media is something humans have been doing without support from AI for as long as we care to recall. See The Daily Mail and other tabloids in the UK’s decades-long campaign of disinformation about the European Union, for an example of this.

The Economist recently provided a helpful infographic of disinformation spread by the British media since the early 90s, sadly including stories published by non-tabloids such as the BBC and The Times.

Source: The Economist

The eyes! The eyes!

Another deep fake demo that crossed my awareness this week was a website called This Person Does Not Exist, which claims to showcase images of human faces that have been generated by an algorithm, something that previously had been hard to achieve.

Hit refresh and another fake face pops up. Obviously helped by foreknowledge, I thought that many of these would bear up to a glance, but still teetered on the edge of the uncanny valley.

The giveaway – when there is one – always seems to be the eyes. Sometimes more obviously than others…

Although other things can go wrong too…

I’m not sure about the provenance of this website. I’m aware that I may be enjoying it because it plays on whatever cognitive biases I have that make me want to believe that machines will never be able to completely fool me. In fact, they probably already have.

Google’s “Yay” vs. Twitter’s nuanced view

A Broadstuff post about the Summly acquisition by Yahoo! looks at the story as a test for how well Google works as a search engine vs. Twitter. 

Now, Google works better than anything out there if you know what it is you want to find, but Twitter, Broadstuff asserts, is where you go to understand what’s really going on… 

Read Google, and you’d barely know anything about Summly because the first 7 pages comprise of press regurgitation and it has utterly failed at telling you anything useful about it….

…But search Twitter, and you get a totally different story. Twitter, despite a reputation for being celebrity and inanity obsessed, is in fact – on the basis of my search anyway, far less so than Google. What is certain is that Twitter gave me a far fuller picture, within the first page I got, and, in this case anyway was the better search engine by far.

The whole media world optimises for Google, it goes on to say, which is making it less useful.

Dronestagram

This, I did not  know:

UK drones have been used in almost 350 attacks in Afghanistan since 2008 (that’s a drone strike every four days), and recently moved to double the size of its fleet of reaper drones.

James Bridle thinks it is something more people should know about and has created an  Instagram profile called Dronestagram where he posts Google Maps satellite images of the places where recent attacks have happened.

API and they know it: Guardian distributes *everything* online

Image: Help yourself (to the Guardian's data
Image: Help yourself (to the Guardian's data)

* Updated *

I’ve also written about Best Buy setting its catalogue content free at the iCrossing Connect blog…

Jeff Jarvis has an excellent post headed APIs: The New Distribution about The Guardian’s decision to distribute everything online.

If you’re even slightly non-technical you may not know what an API is. Basically it’s a way of letting anyone who wants to take Guardian content (headlines, copy, images, video) as it is published and do something different with it.

It makes its content more portable, more shareable, more distributable.

It means The Guardian has taken the limits off of its own content, the limits of what it can think to do with it, and of what can happen on its own site. Feeds from its content will be fed into the most groundbreaking, gamechanging ideas of the next few years (and some duff ones too).

One of Jeff Jarvis’s colleagues describes the move as putting its content “into the fabric of the internet.”

This is a bold move, but one that shows the web literacy of the Guardian Media Group: it understands thefundamentals of being a brand in networks, that it is best served by being in the networks, making itself as useful as possible. It’s just taken the logical next… leap.

This comes at the same time as the BBC is freeing up its news videos to be embedded in other websites.

All well and good – neither organisation is beholden to a quarterly P&L. The Guardian’s a trust and the BBC is a publicly funded (and generously so) corporation. Makes you think that maybe companies that aren’t for profit are the ones who stand the best chance of surviving the gear crunch of adapting to the web. Maybe traditional commercial models aren’t going to be as good at  surviving when it comes to media?

Apart from trusts and public money, the other players in the media mix are the brands. They used to fund the media through advertising mostly, but now will be direct players. How many of them would win in the attention markets by releasing data through APIs like this? Insurance companies have giga-wotsits of useful information. So do publishers, so do pharma companies, so do most people.

If you could, what data from your organisation would put out through an API tomorrow?

Innovation in news: Al-Jazeera Labs

Image: Al Jazeera Labs website
Image: Al Jazeera Labs website

Fascinating to see how Arabic and English news service Al Jazeera is approaching the innovation imperative with its Al Jazeera Labs project.

In the first couple of months of this year the company has rolled out many deals and pilots in interesting areas, according to a great report on Journalism.co.uk. I’m especially intrigued by things like its experiments with Creative Commons licensing of content and use of data visualisation in news stories like the recent war in Gaza.

The map is using both mainstream media reports and what people are saying in social media, via Usahidi, a “platform for crowdsourcing crisis information”. It is designed to help build up a picture of what is happening in a crisis situation – be it a natural disaster or a military conflict – based on what people are saying (by text, blog, Twitter etc.) on the ground.

Image: Al Jazeera's Gaza map on the Ushahidi platform
Image: Al Jazeera's Gaza map on the Ushahidi platform

It’s a very interesting concept, and interesting to see serious attempts to make sense of and filter the rich information – with all the sensible caveats about reliability – that personal content from people involved or near to a crisis situation create.

Here’s a map that has been created from data about the violence in the Democratic Republic of Congo.

Image: An Ushahidi map of incidents in DRC Congo
Image: An Ushahidi map of incidents in DRC Congo

Edge of a riot: Social media, balance and truth in the news

Image: A police line forms toward the end of yesterday's Gaza protest in London (credit: Rich Lewis)

When I was a student in 1994 I was on the front cover of The Indpendent the morning after a riot outside the Houses of Parliament.

The image was of a grimacing, dreadlocked fellow’s grimacing face lunging over the line of police shields.

(No, that wasn’t me…)

The picture spoke a thousand words. It told the whole story. The whole story of a photographer standing the other side of police barricade.

The image looked as if it was taken in the heat of the disturbance. In fact it was a while before anything had happened, when what would become a riot was still a peaceful protest against the Criminal Justice Bill. The man was drunk and on his own. I saw him have a tussle with the cordon of police and – rightly so – being arrested and taken away.

Far from being part of an angry mob there was no one behind him. Well, I was – a few metres back and hence I was in the shot.

Being *in* the protest was a very different experience to being the safer side of the police lines.

After yesterday’s protests in London about Gaza yesterday turned to violence, much of the news coverage is, understandably, about the riot, with few of the images and little of the copy dwelling on the rest of the day of protest. If it bleeds it leads, as they say…

Image: A policeman in riot gear at yesterday's protest (credit: Tyron Francis)

The non-bleeding, peaceful protests get their own coverage in social media. A search for “London protests” filtered by most recent brings images from today’s pro-Israel protests in London, then hundreds of images of yesterday’s March. There are the beginnings of trouble in there (police changing into riot gear as the mood gets uglier, fireworks going off outside the Israeli embassy) and some of the actual violence.

No doubt that in part reflects the priorities of people caught up in the violence (taking part / trying to get away rather than documenting the moment) but perhaps also gives a more proportional balanced view of how the day unfolded. The creativity and passion of the protesters, the diversity of people taking part, the scale of the event are there in the hundreds of photos people have uploaded.

Image: A family on the protest march (credit: Tyron Francis)

The truth is more prosaic, less dramatic, slower than the news cycle. But at a time when churnalism and misinformation is decaying the media’s usefulness as a truthful recorder of events, sometimes social media is where we need to turn for the facts.

: : I went back to the Flickr search as I finished this article and there were many more images of the violence at the end of the day being posted…

There are of course,

For a protester’s-eye view of being on the the march have a look at this:

Gaza protest in London from maryrosecook on Vimeo.

This one follows the news media’s format a little more closely, with the most of it being of the rioting at the end of the day. In big protests like this one, there are often people who are really there with the hop of provoking and tkaing part in trouble, masking their hooliganism as political activism.

People’s News: social media and newsgathering in 2008

From Mumbai to Steve Jobs - 2008's news
From Mumbai to Steve Jobs - social media played a role in the way news was reported in 2008

The staff of NowPublic, the “particpatory news network” as it describes itself, has picked a list of the top 10 stories from 2008 in which social media played a role.

2008’s Top 10 Moments in User-Generated News

1.    Mumbai attacks
2.    Natural disasters: Emergency info
3.    SF Olympic torch relay protests
4.    Obama and “Bittergate”
5.    Protests at Republican Convention
6.    Ushahidi: Crowdsourcing crisis info
7.    CNN’s news wire plans
8.    Mob rule: Mark Zuckerberg at SXSW
9.    Twitter gets student out of Egypt jail
10.  Fake report on Steve Jobs heart attack

As NowPublic’s Rachel Nixon puts it on their blog:

The relationship between producers and consumers of news is changing. What used to be known as “the story” is evolving into something different: fragments of information that don’t come pre-assembled or filtered. With such a rich array of information from so many different sources, it can be confusing without the mechanisms to make sense of it all. We’re in it together to make sense of the story.

Alfred Hermida at Reportr.net says of the list:

Top […] are the Mumbai attacks, a tragic event that demonstrated the value of raw and unfiltered information. It ends with the false report on Steve Jobs heart attack, a salutary tale of the perils of not checking this raw information.

Journalists and the public alike are learning fast about how to get information in raw and filtered forms. Looking forward to seeing how the relationships and processes around gathering and making sense of news evolve in 2009.

Un-filtered news: Twitter, the BBC and Mumbai

Media Guardian carries a timely analysis of some of the discussion of Twitter informing, in some cases becoming part of, the coverage of the terrorist attrocities in Mumbai. 

It picks up on a blog post by Steve Herman, editor of the BBC News website: 

 

As for the Twitter messages we were monitoring, most did not add a great amount of detail to what we knew of events, but among other things they did give a strong sense of what people connected in some way with the story were thinking and seeing. “Appalled at the foolishness of the curious onlookers who are disrupting the NSG operations,” wrote one. “Our soldiers are brave but I feel we could have done better,” said another. There was assessment, reaction and comment there and in blogs. One blogger’s stream of photos on photosharing site Flickr was widely linked to, including by us.

All this helped to build up a rapidly evolving picture of a confusing situation. 

 

Where Twitter added to the understanding of what was happening for a reader of news, was the emotional immediacy. There were voices of people like me, on Twitter, shouting out loud about the horror happening around them. 

It brought Mumbai closer. That’s a good thing, because the whole world needs to feel closer to events like these, the more likely for people to act in small choices and large to fight against religious fundamentalism and zealotry.

 

Image of a peace march by Mumbai blogger Vinu (http://vinu.wordpress.com)
Image of a peace march by Mumbai blogger Vinu (http://vinu.wordpress.com)

 

But in the newsroom at the BBC it also led to a rumour being reported as fact (that the Indian government had asked people to stop using Twitter), which Steve feels was a mistake that should not be repeated: 

 

Should we have checked this before reporting it? Made it clearer that we hadn’t? We certainly would have done if we’d wanted to include it in our news stories (we didn’t) or to carry it without attribution. In one sense, the very fact that this report was circulating online was one small detail of the story that day. But should we have tried to check it and then reported back later, if only to say that we hadn’t found any confirmation? I think in this case we should have, and we’ve learned a lesson. The truth is, we’re still finding out how best to process and relay such information in a fast-moving account like this.

 

More rumour, more noise, more information, more pitfalls: that’s what the continued onward march of social media means for news organisations. 

It’s not a new challenge for the Beeb, as a news organisation that puts accuracy and fact above rumour. Twitter just adds a host of potential sources to the mix during a breaking news story. 

I wrote about this a few years ago in post called Rumour or Raw Data, during the pre-Twitter age if you can remember that, when the then Metropolitan Poolic Commissioner, Sir Ian Blair, said he instinctively turned on Sky News when the first news of the London tube bombings reached him.  

In the heat of the moment, our instinct says I would rather have the unfiltered news, with the risks of inaccuracies and misinformation, than be late to hear. But – and it’s a significant but – you always want the option of a flight to fact: and that fact is usually found on the BBC, 

The BBC is right to resist falling the Sky News of the line. That’s its role: to tell us what the truth is, when it is as sure as it can be what the truth is.

Techmeme turns up the human in its news aggregator mix

Aggregators were my first love, when it comes to news and social media – I’ve always been infatuated with the idea of Digg, lover of Techmeme and I basically see most of the world through my personal aggregator, the ever flexible and accesible Google Reader. 

So it’s nice to see some fervent discussion among bloggers about the best combination. Some love Twitter for the links it brings their way (so do I, sometimes), others eschew the RSS reader for a combo of automated aggregators like Techmeme and Hacker News

What’s sparked the discussion is Techmeme (which has sister aggregator sites Memeorandum (web / tech news), Ballbug (baseball news) and the auto-scurrilacious We Smirch (celeb gossip)) announcing that it will be introducing moe human editorial interventions to keep its list fresh. 

In part this is a response to people trying to game the Techmeme algorithm to give their posts prominence (shame on you). But, as Gabe Riviera, creator of Techmeme explains, it’s also because in lots of other ways “Guess what? Automated news doesn’t quite work“. 

Humans have always edited Techmeme of course, just implicitly. For instance, when a blogger links to a story, the headline might move higher on Techmeme. What’s different now is that an additional human editor will carry out changes explicitly to directly improve the mix of headlines on Techmeme.

I really like the explicit / implicit way of explaining this. Even the great technical marvel that is the Google search engine algorithm is implicitly affected by humans – it is trying to read the clues (links, traffic, words, reputation) that people leave as to which are the best websites on any given keyword.

As Riviera points out – by way of a link to VentureBeat – even Google News has problems in adapting to the mercurial and unpredictable shapes of breaking news. 

Alan Patrick at Broadstuff has an interesting slant on this topic too. Taking a historical analogy, he says that it is early days still for news aggregation:

Long term we suspect bit by bit the human bits of curation will be replaced by better and more intelligent automation. We are in the spinning jenny phase of automated aggregation…. just starting to pick up the threads, as it were :-D 

: : Just read this post by Adam Tinworth, who heads up blogs for Reed Business Information – his take on Techmeme is about the significance for news sites:

“This is a high traffic tech news site – run by one editorial person.”