In 2015, car-and-rocket man Elon Musk joined with influential startup backer Sam Altman to put artificial intelligence on a new, more open course. They cofounded a research institute called OpenAI to make new AI discoveries and give them away for the common good. Now, the institute's researchers are sufficiently worried by something they built that they won't release it to the public. The AI system that gave its creators pause was designed to learn the patterns of language. It does that very well--scoring better on some reading-comprehension tests than any other automated system.

Better known as a supplier of facial recognition software used by the Chinese government, an AI-startup that is backed by Alibaba has developed software that can identify dogs by their noses. No, it isn't April 1st; the facial recognition software developed by Megvii really can identify one dog from another by using nasal biometrics. KrAsia news reports that the company has developed the software on the basis that dogs have unique nose prints. Dr. David Dorman, a professor of toxicology, has previously said that: "Like human fingerprints, each dog has a unique nose print. Some kennel clubs have used dog nose prints for identification."

A team at Flinders University in South Australia has developed a new vaccine believed to be the first human drug in the world to be completely designed by artificial intelligence (AI). While drugs have been designed using computers before, this vaccine went one step further being independently created by an AI program called SAM (Search Algorithm for Ligands). Flinders University Professor Nikolai Petrovsky who led the development told Business Insider Australia its name is derived from what it was tasked to do: search the universe for all conceivable compounds to find a good human drug (also called a ligand). "We had to teach the AI program on a set of compounds that are known to activate the human immune system, and a set of compounds that don't work. The job of the AI was then to work out for itself what distinguished a drug that worked from one that doesn't," Petrovsky said, who is also the Research Director of Australian biotechnology company Vaxine.

And one in which Matthew McConaughey took the Leo role in Titanic. And Saved by the Bell's Tiffani Thiessen played Rachel in Friends. The entertainment industry isn't exactly short on "what if?" scenarios in which actors came close to, but were ultimately passed over, playing iconic roles. For more than 99% of movie history, fans have been able to do little more than squirrel away this trivia for use in pop quizzes. That is until the arrival of deepfakes.

When it comes to exploring far off planets, robots need to be able to tackle all sorts of challenges, so NASA has been working on a series of climbing robots to take on different tasks in inhospitable environments. First up is LEMUR (Limbed Excursion Mechanical Utility Robot) which can climb rock walls using hundreds of fishhooks in its fingers. It uses A.I. to navigate around obstacles that it cannot climb, and is one of NASA's first generation of climbing robots. It was developed to perform repair tasks aboard the International Space Station, and below you can see it in a field test in Death Valley, California. Then there's the somewhat terrifying-looking Ice Worm, which was adapted from one of LEMUR's limbs.

The Myers-Briggs Type Indicator (MBTI) is a popular personality metric that uses four dichotomies as indicators of personality traits. This paper examines the use of pre-trained language models to predict MBTI personality types based on scraped labeled texts. The proposed model reaches an accuracy of $0.47$ for correctly predicting all 4 types and $0.86$ for correctly predicting at least 2 types. Furthermore, we investigate the possible uses of a fine-tuned BERT model for personality-specific language generation. This is a task essential for both modern psychology and for intelligent empathetic systems.

Risk adjustment has become an increasingly important tool in healthcare. It has been extensively applied to payment adjustment for health plans to reflect the expected cost of providing coverage for members. Risk adjustment models are typically estimated using linear regression, which does not fully exploit the information in claims data. Moreover, the development of such linear regression models requires substantial domain expert knowledge and computational effort for data preprocessing. In this paper, we propose a novel approach for risk adjustment that uses semantic embeddings to represent patient medical histories. Embeddings efficiently represent medical concepts learned from diagnostic, procedure, and prescription codes in patients' medical histories. This approach substantially reduces the need for feature engineering. Our results show that models using embeddings had better performance than a commercial risk adjustment model on the task of prospective risk score prediction.

In distributed reinforcement learning, it is common to exchange the experience memory of each agent and thereby collectively train their local models. The experience memory, however, contains all the preceding state observations and their corresponding policies of the host agent, which may violate the privacy of the agent. To avoid this problem, in this work, we propose a privacy-preserving distributed reinforcement learning (RL) framework, termed federated reinforcement distillation (FRD). The key idea is to exchange a proxy experience memory comprising a pre-arranged set of states and time-averaged policies, thereby preserving the privacy of actual experiences. Based on an advantage actor-critic RL architecture, we numerically evaluate the effectiveness of FRD and investigate how the performance of FRD is affected by the proxy memory structure and different memory exchanging rules.

We propose a new batch mode active learning algorithm designed for neural networks and large query batch sizes. The method, Discriminative Active Learning (DAL), poses active learning as a binary classification task, attempting to choose examples to label in such a way as to make the labeled set and the unlabeled pool indistinguishable. Experimenting on image classification tasks, we empirically show our method to be on par with state of the art methods in medium and large query batch sizes, while being simple to implement and also extend to other domains besides classification tasks. Our experiments also show that none of the state of the art methods of today are clearly better than uncertainty sampling when the batch size is relatively large, negating some of the reported results in the recent literature.

Twin-to-twin transfusion syndrome treatment requires fetoscopic laser photocoagulation of placental vascular anastomoses to regulate blood flow to both fetuses. Limited field-of-view (FoV) and low visual quality during fetoscopy make it challenging to identify all vascular connections. Mosaicking can align multiple overlapping images to generate an image with increased FoV, however, existing techniques apply poorly to fetoscopy due to the low visual quality, texture paucity, and hence fail in longer sequences due to the drift accumulated over time. Deep learning techniques can facilitate in overcoming these challenges. Therefore, we present a new generalized Deep Sequential Mosaicking (DSM) framework for fetoscopic videos captured from different settings such as simulation, phantom, and real environments. DSM extends an existing deep image-based homography model to sequential data by proposing controlled data augmentation and outlier rejection methods. Unlike existing methods, DSM can handle visual variations due to specular highlights and reflection across adjacent frames, hence reducing the accumulated drift. We perform experimental validation and comparison using 5 diverse fetoscopic videos to demonstrate the robustness of our framework.