The Austin-based company revealed research at SXSW that analyzed over 2,000 film scripts and 25,000 characters from between 1930 and 2018. Using machine learning, sentiment analysis and natural language processing, StoryFit's approach demonstrates new ways to break apart narratives and character personality. The company says they ask over 10,000 queries of every script, focusing on what characters say and how they say it. Analyzing the "Big 5" personality traits, they found that 80 percent of female characters measured scored over 85 percent on agreeableness, while only 27 percent of men scored that highly. Characters like Rey from Star Wars: The Force Awakens, who score low on agreeableness, are outliers among female characters.

CREATIVITY: IT'S AT THE HEART OF WHO WE HUMANS ARE… WE HUMANS ARE SPECIAL, RIGHT? Can a robot write a symphony? Can a robot turn a canvas into a beautiful masterpiece? OVER SOME 40,000 YEARS, HUMAN CREATIVITY HAS EXPLODED – FROM DRAWINGS ON CAVE WALLS THROUGH THE GREAT ART OF CENTURIES TO COME…. NOW, SCIENTISTS -- AND ARTISTS –ARE ASKING CAN A ROBOT TRULY IMAGINE AN ORIGINAL MASTERWORK? COMPUTATIONAL CREATIVITY IS LEADING US TO ASK NEW QUESTIONS ABOUT HUMAN CREATIVITY. IS THIS ESSENTIAL HUMAN TRAIT TRULY UNIQUE? WILL ARTIFICIAL INTELLIGENCE BE A COMPETITOR? OR CAN IT BE A COLLABORATOR, HELPING US TOWARD STILL UNIMAGINED CREATIONS? SCHAEFER: My first guest is a member of Google Brain's Magenta team. He is currently working on neural network models of sound and music and recently produced a synthesizer that designed its own sounds. SCHAEFER: Also with us, is an Assistant professor at the University of Illinois at Urbana Champaign in the Dept. of Electrical and Computer Engineering. He focuses on several surprising creative domains including the culinary arts and fashion and the theoretical foundations of creativity. SCHAEFER: Also with us is an Associate Professor of psychological and brain science at Dartmouth College. He's interested in the neural basis of imagination and in the evolution of human creativity. A former research fellow at MIT's Media lab and artist in residence at Google, please welcome Sougwen Chung.

I'm currently implementing an A3C agent in Tensorflow (Asynchronous Advantage Actor Critic) that plays doom (using vizdoom) and I was thinking about if there is a difference between using CNNs or Capsnets (Capsule Networks), Recently there was a big breakthrough in computer vision with these Capsnets. I know that Capsnets, instead of Convnets, handle the spatial relationship of the features and detecting rotated objects. As a consequence, I wondered if there is an advantage to use Capsnets in a Deep Reinforcement Learning agent?

Reuters is building a tool with artificial intelligence to help journalists in analyzing data, suggest story ideas, and even write more sentences. The tool aims not to replace reporters but instead augment them with a digital data scientist-cum-copywriting assistant. Called Lynx Insight, it has been used by many journalists and will now be rolled out across all the classrooms of Reuters. Reg Chua, executive editor of editorial operations, data and innovation at Reuters, says the aim is to divvy up editorial work into what machines do best and what editorial human staff excels at (such as asking questions, judging importance, understanding context and -- presumably -- drinking excessive amounts of coffee). That differs from previous editorial tech efforts that sought to train an AI to write entire stories in the form of snippets about local sports teams or earthquake warnings.

But I am stuck with section 4(Convnet Visualization) and having a hard time understanding the visualizations shown in Figure 2. Can anyone explain these portions to me?

It used to be you could spend an afternoon drumming up a home movie with your little sister, soundtrack it with your favorite mixtape cuts, and upload it to the Internet for sharing without a care. What's an amateur-at-best musician to do? I may have marched on a collegiate snare line (and therefore understand rhythm, phrasings, and tempo), but my ability to create melody probably stopped with middle school recorder lessons. Luckily, we musically challenged filmmakers and podcasters now have robots. A number of high-profile AI composition initiatives have surfaced in recent years--perhaps most notably, Sony's Flow Machine released its debut album in January--and slowly but surely these tools are moving from the research labs and professional production studios into publicly available spaces.

Before any build log, here's a video: That was a Rubik's cube being solved in 0.38 seconds. The time is from the moment the keypress is registered on the computer, to when the last face is flipped. It includes image capture and computation time, as well as actually moving the cube. The motion time is 335 ms, and the remaining time image acquisition and computation. For reference, the current world record is/was 0.637 seconds.

Your "speed things up by compressing them" idea (and, yes, it makes sense and is the basis of a lot of techniques in information retrieval) can also work without the autoencoder. Siamese networks process separate inputs through separate paths where each of the distinct paths share their parameters. What's stopping you from, for all of your repository entries, precomputing the forward pass up to the point where the paths of the network merge and storing the result instead of the raw input? This would mean that at lookup time, you'd only have to do a full forward pass for the query image and then substitute in the precomputed representation for the candidates. All this can be done no matter how the network was trained -- it doesn't rely on training a separate autoencoder as a preprocessor.

I hope you all had a wonderful Christmas and a spectacular start to the New Year! Yes, 2017 should have received a speeding ticket, since it went by too fast! And, I'm sure, 2018 will similarly speed along at an unsuitable pace – apparently time flies when you're having fun, or is it just old age? So, it's 2018, and I thought in this month's column I would steer away from the traditional diet of "Here are my predictions for 2018." To be honest, I simply couldn't swallow writing a piece of this ilk, as I'm sure plenty of others will have their own unique perspective.

We better hope that when the artificial intelligence apocalypse finally descends on us, it'll be something like "Westworld." That's the word from show co-creator Jonathan Nolan, who spoke Saturday at the South by Southwest Conference during a panel about the hit series. "I think we'd be lucky if this was the AI apocalypse, if it was this attractive and charming," Nolan said, noting that while folks tend to think it'll take a super AI to overthrow humans, bots are already manipulating social media users. The implications of AI comprised just some of the discussion during the panel, which in addition to Nolan featured co-creator Lisa Joy, a cast member or two, and a rocket-fast cameo by tech-minded Renaissance man Elon Musk, founder of SpaceX, Tesla and the Boring Company. After a preamble from Nolan on the aspirational qualities of going to space, Musk walked on stage in the last few minutes to show off a short video by Nolan and Joy. The clip focused on the recent launch of SpaceX's Falcon Heavy rocket, and on Starman, the dummy at the wheel of a Tesla SpaceX Roadster carried into space by the rocket.