There has been a large resurgence of interest in generative models recently (see this blog post by OpenAI for example). These are models that can learn to create data that is similar to data that we give them. The intuition behind this is that if we can get a model to write high-quality news articles for example, then it must have also learned a lot about news articles in general. Or in other words, the model should also have a good internal representation of news articles. We can then hopefully use this representation to help us with other related tasks, such as classifying news articles by topic.

On Monday, Google announced plans to launch a new peer review journal and "ecosystem" for machine learning. Scientists need to develop notations, analogies, visualizations, and explanations of ideas. It's deeply tied to the heart of science. "That's why, in collaboration with OpenAI, DeepMind, YC Research, and others, we're excited to announce the launch of Distill, a new open science journal and ecosystem supporting human understanding of machine learning. Distill is an independent organization, dedicated to fostering a new segment of the research community. "Modern web technology gives us powerful new tools for expressing this human dimension of science.

He will be discussing the topic at this year's edition of South by South West on March 14th in Austin, Texas. For EPFL theoretical biologist Marcel Salathé, the answer is invariably yes. To him, a more fundamental question that needs to be addressed is who owns that artificial intelligence? "We have to hold AI accountable, and the only way to do this is to verify it for biases and make sure there is no deliberate misinformation," says Salathé. "This is not possible if the AI is privatized." So what exactly is AI? It is generally regarded as "intelligence exhibited by machines".

There is growing interest in applying machine learning methods to Electronic Medical Records (EMR). Across different institutions, however, EMR quality can vary widely. This work investigated the impact of this disparity on the performance of three advanced machine learning algorithms: logistic regression, multilayer perceptron, and recurrent neural network. The EMR disparity was emulated using different permutations of the EMR collected at Children's Hospital Los Angeles (CHLA) Pediatric Intensive Care Unit (PICU) and Cardiothoracic Intensive Care Unit (CTICU). The algorithms were trained using patients from the PICU to predict in-ICU mortality for patients in a held out set of PICU and CTICU patients. The disparate patient populations between the PICU and CTICU provide an estimate of generalization errors across different ICUs. We quantified and evaluated the generalization of these algorithms on varying EMR size, input types, and fidelity of data.

Deep learning has been used to transform artificial intelligence (AI) development, whether it is from beating players in games like Go or poker to improving self-driving AI. But perhaps the most important changes for most of us is how AI advances and machine learning are affecting healthcare. In January, a medical startup won FDA approval for an AI-assisted cardiac imaging system called Arterys, and AI is playing vital roles in other health fields such as fighting cancer and aging. NVIDIA boasts that with deep learning, "AI can help doctors make faster, more accurate diagnoses. It can predict the risk of a disease in time to prevent it."

New research from OpenAI and UC Berkeley has created A.I. agents that can form and use their own new language, without instruction, whenever they need to. The languages are systematic and roughly grammatical, and even include aspects of non-verbal communication like body language! It all makes for an incredible glimpse into how (and why) language may have arisen during biological evolution, and it shows the nuanced insight we can derive from modern learning agents. Like so many studies that set out to elicit a specific A.I. behavior, this one began by creating a rough metaphor for real life. The experiment sets its A.I. agents in a simulated physical world containing landmarks at fixed positions, and then gives them the ability to roam freely within this two-dimensional space. The agents were then given a goal, usually to send another agent to a specific place in the world, and a set of nonsense symbols each could "say" aloud so the others could "hear" it.

Artificial intelligence (AI) expert Andrew Ng has announced that he is resigning from his role as chief scientist at Chinese search engine giant Baidu after nearly three years in the job. Ng, who announced his departure in a blog post on Wednesday, does not currently have another job lined up, although he's likely to be in high demand. "I will be resigning from Baidu, where I have been leading the company's AI Group," wrote Ng in the Medium blog post. "After Baidu, I am excited to continue working toward the AI transformation of our society and the use of AI to make life better for everyone." Along with the likes of DeepMind, Google, Microsoft and Facebook, Baidu is often seen as one of the world leaders when it comes to AI research.

What is the best way to start learning machine learning and deep learning without taking any online courses? Let me first start off by saying that there is no single "best way" to learn machine learning, and you should find a system that works well for you. Some people prefer the structure of courses, others like reading books at their own pace, and some want to dive right into code. I started with Andrew Ng's Machine Learning Coursera course in 2012, knowing almost zero linear algebra and nothing about statistics or machine learning. Note that although the class covered neural networks, it was not a course on Deep Learning.

In 1971, Terry Winograd wrote the SHRDLU program while completing his PhD at MIT. SHRDLU features a world of toy blocks where the computer translates human commands into physical actions, such as "move the red pyramid next to the blue cube." To succeed in such tasks, the computer must build up semantic knowledge iteratively, a process Winograd discovered was brittle and limited. The rise of chatbots and voice activated technologies has renewed fervor in natural language processing (NLP) and natural language understanding (NLU) techniques that can produce satisfying human-computer dialogs. Unfortunately, academic breakthroughs have not yet translated to improved user experiences, with Gizmodo writer Darren Orf declaring Messenger chatbots "frustrating and useless" and Facebook admitting a 70% failure rate for their highly anticipated conversational assistant M. Nevertheless, researchers forge ahead with new plans of attack, occasionally revisiting the same tactics and principles Winograd tried in the 70s. OpenAI recently leveraged reinforcement learning to teach to agents to design their own language by "dropping them into a set of simple worlds, giving them the ability to communicate, and then giving them goals that can be best achieved by communicating with other agents."

Summary: Some observations about new major trends and directions in data science drawn from the Strata Hadoop conference in San Jose last week. This is always exciting, enervating, and exhausting but it remains the single best place to pick up on what's changing in our profession. This conference is on a world tour with four more stops before repeating next year. The New York show is supposed to be a little bigger (hard to imagine) but the San Jose show is closest to our intellectual birthplace. After all this is the place where to call yourself a nerd would be regarded as a humble brag.