Twitter HQ. Photo: Kevin Krejci via Flickr.
Show Hide image

Twitter's new porn-spotting robot moderators

The social networking site has introduced new artificial intelligence systems that can spot and delete sexual and violent images – and spare human moderators in the process. 

Under the compelling headline “The labourers who keep dick pics and beheadings out of your Facebook feed”, journalist Adrien Chen delved last year into the little-known world of social media’s content moderators. These thousands of workers, most based in Asia, trawl through social networking sites in order to delete or flag offensive content. In the process, they are exposed to the very worst the internet has to offer – beheadings, violent pornography, images of abuse – all for wages as low as $300 a month.

But this month, Twitter has taken a first step towards automating this process, and thus sparing a huge unseen workforce from their daily bombardment of horrors. Almost exactly a year ago, Twitter bought start-up Madbits, which offers, in the words of its co-founders, a “visual intelligence technology that automatically understands, organises and extracts relevant information from raw media”. 

At the time, tech websites speculated that the Madbits would be used to develop facial recognition or tagging on Twitter photos. But in fact, the start-up’s first task was very different: it was instructed by Alex Roetter, Twitter’s head of engineering, to build a system which could find and filter out offensive images, defined by the company as "not safe for work". 

This month, Wired reported that these artificial intelligence (AI) moderators are now up and running. Roetter claims the new moderator-bots can filter out 99 per cent of offensive imagery. They also tend to incorrectly identify about 7 per cent of acceptable images as offensive – but, the company reasons, better safe than sorry. 

Like other artificial intelligence robots, the moderator “learns” how to spot offensive imagery by analysing reams of pornography and gore, and then applies its knowledge of the content and patterns to new material. Over time, the system continues to learn, and get even better at spotting NSFW images. Soon, systems like these could replace content moderation farms altogether.

In most cases like these, it's worth remembering those whose jobs might be lost as the robots advance, especially in developing countries – but considering the psychological damage brought on by endless exposure to violent images, we can only hope Twitter and sites like it can offer less distressing moderation jobs (and higher salaries) to these workers instead. 

Barbara Speed is a technology and digital culture writer at the New Statesman and a staff writer at CityMetric.

Davide Restivo at Wikimedia Commons
Show Hide image

Scientists have finally said it: alcohol causes cancer

Enough of "linked" and "attributable": a new paper concludes that alcohol directly causes seven types of cancer.

I don't blame you if you switch off completely at the words "causes cancer". If you pay attention to certain publications, everything from sunbeds, to fish, to not getting enough sun, can all cause cancer. But this time, it's worth listening.

The journal Addiction has published a paper that makes a simple, yet startling, claim: 

"Evidence can support the judgement that alcohol causes cancer of the oropharynx [part of the throat], larynx, oesophagus, liver, colon, rectum and [female] breast"

So what's especially significant about this? 

First, scientists, unlike journalists, are very wary of the word "causes". It's hard to ever prove that one action directly led to another, rather than that both happened to occur within the same scenario. And yet Jennie Connor, author of the paper and professor in the Preventive and Social Medicine department at the University of Otago, New Zealand, has taken the leap.

Second, alcohol not only causes cancer of one kind – the evidence supports the claim that it causes cancer at seven different sites in our bodies. There was weaker evidence that it may also cause skin, prostate and pancreatic cancer, while the link between mouth cancers and alcohol consumption was the strongest. 

What did we know about alcohol and cancer before?

Many, many studies have "linked" cancer to alcohol, or argued that some cases may be "attributable" to alcohol consumption. 

This paper loooks back over a decade's worth of research into alcohol and cancer, and Connor concludes that all this evidence, taken together, proves that alcohol "increases the incidence of [cancer] in the population".

However, as Connor notes in her paper, "alcohol’s causal role is perceived to be more complex than tobacco's", partly because we still don't know exactly how alcohol causes cancer at these sites. Yet she argues that the evidence alone is enough to prove the cause, even if we don't know exactly how the "biologial mechanisms" work. 

Does this mean that drinking = cancer, then?

No. A causal link doesn't mean one thing always leads to the other. Also, cancer in these seven sites was shown to have what's called a "dose-response" relationship, which means the more you drink, the more you increase your chances of cancer.

On the bright side, scientists have also found that if you stop drinking altogether, you can reduce your chances back down again.

Are moderate drinkers off the hook?

Nope. Rather devastatingly, Connor notes that moderate drinkers bear a "considerable" portion of the cancer risk, and that targeting only heavy drinkers with alcohol risk reduction campaigns would have "limited" impact. 

What does this mean for public health? 

This is the tricky bit. In the paper, Connor points out that, given what we know about lung cancer and tobacco, the general advice is simply not to smoke. Now, a strong link proven over years of research may suggest the same about drinking, an activity society views as a bit risky but generally harmless.

Yet in 2012, it's estimated that alcohol-attributable cancers killed half a million people, which made up 5.8 per cent of cancer deaths worldwide. As we better understand the links between the two, it's possible that this proportion may turn out to be a lot higher. 

As she was doing the research, Connor commented:

"We've grown up with thinking cancer is very mysterious, we don't know what causes it and it's frightening, so to think that something as ordinary as drinking is associated with cancer I think is quite difficult."

What do we do now?

Drink less. The one semi-silver lining in the study is that the quantity of alcohol you consume has a real bearing on your risk of developing these cancers. 

On a wider scale, it looks like we need to recalibrate society's perspective on drinking. Drug campaigners have long pointed out that alcohol, while legal, is one of the most toxic and harmful drugs available  an argument that this study will bolster.

In January, England's chief medical officer Sally Davies introduced some of the strictest guidelines on alcohol consumption in the world, and later shocked a parliamentary hearing by saying that drinking could cause breast cancer.

"I would like people to take their choice knowing the issues," she told the hearing, "And do as I do when I reach for my glass of wine and think... do I want to raise my risk of breast cancer?"

Now, it's beginning to look like she was ahead of the curve. 

Barbara Speed is a technology and digital culture writer at the New Statesman and a staff writer at CityMetric.