Technology

The EU-funded ‘What-if Machine’ (WHIM) project not only generates fictional storylines but also judges their potential usefulness and appeal. It represents a major advance in the field of computational creativity

Science rarely looks at the whimsical, but that is changing as a result of the aptly named WHIM project. The ambitious project is building a software system able to invent and evaluate fictional ideas.
‘WHIM is an antidote to mainstream artificial intelligence which is obsessed with reality,’ says Simon Colton, project coordinator and professor in computational creativity at Goldsmiths College, University of London. ‘We’re among the first to apply artificial intelligence to fiction.’
World’s first fictional ideation machine

The project acronym stands for the What-If Machine. It is also the name of the world’s first fictional ‘ideation’ (creative process of generating, developing, and communicating new ideas) software, developed within the project. The software generates fictional mini-narratives or storylines, using natural language processing techniques and a database of facts mined from the web (as a repository of ‘true’ facts). The software then inverts or twists the facts to create ‘what-ifs’. The result is often incongruous, ‘What if there was a woman who woke up in an alley as a cat, but could still ride a bicycle?’

Can computers judge creativity?

WHIM is more than just an idea-generating machine. The software also seeks to assess the potential for use or quality of the ideas generated. Since the ideas generated are ultimately destined for human consumption, direct human input was asked for in crowd sourcing experiments. For example, WHIM researchers asked people whether they thought the ‘what-ifs’ were novel and had good narrative potential, and also asked them to leave general feedback. Through machine learning techniques, devised by researchers at the Jozef Stefan Institute in Ljubljana, the system gradually gains a more refined understanding of people’s preferences.
‘One may argue that fiction is subjective, but there are patterns,’ says Professor Colton. ‘If 99% of people think a comedian is funny, then we could say that comedian is funny, at least in the perception of most people.’

Just the beginning

Generating fictional mini-narratives is just one aspect of the project. Researchers at the Universidad Complutense Madrid are expanding the mini-narratives into full narratives that could be more suitable for the complete plot of a film, for example. Meanwhile, researchers at the University College in Dublin are trying to teach computers to produce metaphorical insights and ironies by inverting and contrasting stereotypes harvested from the web, while researchers from the University of Cambridge are looking into web mining for ideation purposes. All of this work should lead to better and more complete fictional ideas.

More than a whim

While the fictional ideas generated may be whimsical, WHIM is based on solid science. It is part of the emerging field of computational creativity, a fascinating interdisciplinary discipline located at the intersection of artificial intelligence, cognitive psychology, philosophy, and the arts.

WHIM may have applications in multiple domains. In one initiative there are plans to turn the narratives into video games. Another major initiative involves the computational design of a musical theatre production: the storyline, sets and music. The entire process is being filmed for a documentary.

WHIM could also be applied in areas beyond the arts. For example, it could be used by moderators at scientific conferences to ask probing ‘what-if’ questions to panellists in order to explore different hypotheses or scenarios.

References:http://phys.org/

A trio of researchers at the University of Southern California has uncovered a social-network illusion that might explain why some things become popular in cyberspace while others do not. Kristina Lerman, Xiaoran Yan and Xin-Zeng Wu have written a paper describing the illusion and how it works and have posted it on the preprint server arXiv.

Social networks are not new of course, they have been going on for thousands of years, if not longer—what is new is the venue and size. As they have grown online, scientists have begun studying them in earnest and have found some interesting things, one of which is the friendship paradox—where any given person’s “friends” will have more friends than they have. This illusion is created by the slewing of the average by one or more friends that have a lot of friends. And it is not restricted to just friending sites, studies have shown that for the average Twitter user, their friends will Tweet more than they do, and again, it is an illusion that comes about due to slewing by just a few other users. In this new effort, the researchers have found a similar illusion, where ideas, photos or other information can appear to be much more popular than they really ar

The illusion comes about, the team explains, due to just a few nodes (people) having links to a lot of others—they provide an illustration of two views of a simple 14-node network, the only difference between them is that different nodes have been colored red. In one view, nodes with multiple links have been colored, in the other, those with just a few links have been colored. The researchers then suggest the viewer consider the perspective of nodes that are not colored, under the first scenario—any of them will see whatever message is being given by one of the more popular (red) nodes—and that is where the illusion occurs. In the real world, it appears possible that such networks would allow something to seem much more popular than it really is, because it is being disseminated by just a few well-connected nodes, whether it is a video of a cat doing something stupid, or a minority opinion about a well known topic.

Abstract

Social behaviors are often contagious, spreading through a population as individuals imitate the decisions and choices of others. A variety of global phenomena, from innovation adoption to the emergence of social norms and political movements, arise as a result of people following a simple local rule, such as copy what others are doing. However, individuals often lack global knowledge of the behaviors of others and must estimate them from the observations of their friends’ behaviors. In some cases, the structure of the underlying social network can dramatically skew an individual’s local observations, making a behavior appear far more common locally than it is globally. We trace the origins of this phenomenon, which we call “the majority illusion,” to the friendship paradox in social networks. As a result of this paradox, a behavior that is globally rare may be systematically overrepresented in the local neighborhoods of many people, i.e., among their friends. Thus, the “majority illusion” may facilitate the spread of social contagions in networks and also explain why systematic biases in social perceptions, for example, of risky behavior, arise. Using synthetic and real-world networks, we explore how the “majority illusion” depends on network structure and develop a statistical model to calculate its magnitude in a network.

References:http://phys.org/

An algorithm has been designed to tell if somebody in a color photo is naked. Isitnude.com launched earlier this month; its demo page invites you to try it out to test its power in nudity detection. You can choose from a selection of images at the bottom of the page, including pics of Vladimir Putin on horseback and Tiger Woods in golf mode. We tried it out, dragging and dropping a picture of Woods over into the box and the message promptly said “Not nude-G.” “You can probably post this.”

Other notes on the page include, “We apologize if we didn’t get it right, we are improving every day.” “Please note that we cannot detect black and white images.”
(Algorithmia does not retain images.)

The company behind this effort, Algorithmia, was founded in 2013 to advance algorithm development and use. “As developers ourselves we believe that given the right tools the possibilities for innovation and discovery are limitless.”

They said they are building “what we believe to be the next era of programming: a collaborative, always live and community driven approach to making the machines that we interact with better.” The community driven API exposes “the collective knowledge of algorithm developers across the globe.”
“We’re building a community around state-of-the-art algorithm development. Users can create, share, and build on other algorithms and then instantly make them available as a web service.”

Lucy Black in I Programmer noted the use advantage. “The idea behind Algorithmia is that where an algorithm already exists you don’t need to code your own, instead you can simply paste in its functionality using its cloud-based API.”

In his story about Algorithmia and the demo, Brian Barrett in Wired said, “His company is an algorithmic clearing house, taking computational solutions from academia and beyond and offering them to the world at large for a fee.”

Why the interest in detecting nudity in photographs?

“A customer came to us trying to run a site that needs to be kid-friendly,” said Algorithmia CTO Kenny Daniel in Wired. The customer wanted the ability to screen images with some confidence that they would not be pornographic. Daniel said, “Anybody who’s trying to run a community but wants to filter out objectionable content, or keep it kid-friendly, could benefit from this same algorithm.”

In the company blog this month, they also discussed the rationale in enabling artificial intelligence to detect nudity.

“If there’s one thing the internet is good for, it’s racy material,” said the blog. This is a headache for a number of reasons, including a) it tends to show up where you’d rather it wouldn’t, like forums, social media, etc. and b) while humans generally know it when we see it, computers don’t so much. We here at Algorithmia decided to give this one a shot.”

To give it a shot, they turned to various sources. For one, the result is based on an algorithm by Hideo Hattori and on a paper authored by Rigan Ap-apid, De La Salle University. In the latter’s paper, “An Algorithm for Nudity Detection,” he presented an algorithm for detecting nudity in color images.

He said, “A skin color distribution model based on the RGB, Normalized RGB, and HSV color spaces is constructed using correlation and linear regression. The skin color model is used to identify and locate skin regions in an image. These regions are analyzed for clues indicating nudity or non-nudity such as their sizes and relative distances from each other. Based on these clues and the percentage of skin in the image, an image is classified nude or non-nude.”

Meanwhile, plenty of images with lots of skin are “perfectly innocent,” said the blog. “You might say that leaning too much on just color leaves the method, well, tone-deaf. To do better, you need to combine skin detection with other tricks.”

Brian Barrett in Wired said “To help weed out false positives, Algorithmia added a few layers of intelligence.”

The blog stated that “Our contribution to this problem is to detect other features in the image and using these to make the previous method more fine-grained.”
To come up with the algorithm, they turned to the book Human Computer Interaction Using Hand Gestures by Prashan Premaratne, OpenCV’s nose detection algorithm and face detection algorithm.

As I Programmer said, the algorithm is still a work in progress. They are still interested in further improvements. “There are countless techniques that can be used in place of or combined with the ones we’ve used to make an even better solution. For instance, you could train a convolutional neural network on problematic images,” they said.

References:http://phys.org/

UNPUTDOWNABLE! This go-to accolade for book reviewers takes on a new meaning this week, when some authors will be paid by the number of pages read. But is it a good thing?

On 1 July, Amazon introduced a payment plan in which authors enrolled in its Kindle Direct Publishing Select scheme are paid royalties according to the number of pages read when their ebooks are borrowed from ebook “library” Kindle Unlimited. But critics claim this pay-per-page system could change the type of books written, favouring easy-to-read page-turners over difficult, literary works.

It is the latest example of firms and artists struggling to find a balance when it comes to digital distribution. Last week, Taylor Swift persuaded Apple Music to pay for songs played during its free three-month trial period, which it hadn’t intended to do. And Steam, the largest online game store, has revealed that the vast majority of games on its site sell fewer than 32,000 copies – many at a reduced price – which isn’t enough for games studios to survive.
This wasn’t how it was meant to be. “Digital distribution began with this real utopian vision of change,” says James Allen-Robertson at the University of Essex in Colchester, UK. The internet was hailed as a means for artists to find their audience and audiences their artist. But that hasn’t happened. “All these platforms have opened up access to creators, but now the markets are so saturated it’s hard to be found,” he says.

This is having negative effects on the creators. Belgium-based games studio Tale of Tales announced that it would no longer be making games for commercial release after its latest title, Sunset, sold just 4000 copies, despite getting good reviews.

Part of the problem is that the internet is dominated by just a few platforms, like Amazon, Spotify and Steam. Also, instead of searching for new music or books ourselves, we tend to follow what is recommended automatically. But algorithms tend to recommend things that are alike. “It forces a homogenisation of content,” says Auriea Harvey, one of the pair behind Tale of Tales.

In response, a few artists game the system. Music streaming services like Spotify pay a small fee for each play of a song. Some people now release hundreds of albums of mediocre music on Spotify, betting that the economy of scale will result in at least one or two plays – and thus fees. Last week, another group launched Eternify, an app that will play the first 30 seconds of a Spotify song – the minimum needed to qualify for payment – on repeat. Another band launched Sleepify, an album of silence, and encouraged fans to leave it playing to generate royalties. Spotify has now removed both apps.

Similarly, Amazon’s payment model is a reaction to authors gaming their previous system – in which fees were paid if 10 per cent of a book was read – by dividing novels up into serial form.

The tussle over digital distribution will continue to change the works produced. It may not pay to create songs with a slow build-up, for example, as listeners on Spotify may skip ahead before the 30-second payment mark is reached. “The past has shown that the mode of delivery does have a significant impact on the way a thing is expressed,” says Allen-Robertson.

References:http://www.newscientist.com/