Ai

How Waze predicts carpools with Google Cloud’s AI Platform

Waze’s mission is to eliminate traffic and we believe our carpool feature is a cornerstone that will help us achieve it. In our carpool apps, a rider (or a driver) is presented with a list of users that are relevant for their commute (see below). From there, the rider or the driver can initiate an offer to carpool, and if the other side accepts it, it’s a match and a carpool is born.

OpenAI releases powerful text generator

The laboratory, founded by Elon Musk and recently supported by a $1 billion grant from Microsoft, has designed text generators that create readable passages virtually indistinguishable from those written by humans. OpenAI’s machine learning approach scrapes massive amounts of data from the web and analyzes it for statistical patterns that allow it to realistically predict what letters or words will likely be written next. When users feed a word or phrase or longer text snippets into the generator, it expands on the words with convincingly humanlike text.

Artificial brains may need sleep too

No one can say whether androids will dream of electric sheep, but they will almost certainly need periods of rest that offer benefits similar to those that sleep provides to living brains, according to new research from Los Alamos National Laboratory. Watkins and her research team found that the network simulations became unstable after continuous periods of unsupervised learning. When they exposed the networks to states that are analogous to the waves that living brains experience during sleep, stability was restored.

5 Essential Papers on AI Training Data

Many data scientists claim that around80% of their time is spent on data preprocessing, and for good reasons, as collecting, annotating, and formatting data are crucial tasks in machine learning. This article will help you understand the importance of these tasks, as well as learn methods and tips from other researchers. Below, we will highlight academic papers from reputable universities and research teams on various training data topics.

Our weird behavior during the pandemic is messing with AI models

IBM will no longer offer, develop, or research facial recognition technology

Train ALBERT for natural language processing with TensorFlow on Amazon SageMaker

Top 10 Best FREE Artificial Intelligence Courses

Most of the Machine Learning, Deep Learning, Computer Vision, NLP job positions, or in general every Artificial Intelligence (AI) job position requires you to have at least a bachelor’s degree in Computer Science, Electrical Engineering, or some similar field. If your degree comes from some of the world’s best universities than your chances might be higher in beating the competition on your job interview. But looking realistically, not most of the people can afford to go to the top universities in the world simply because not most of us are geniuses and don’t have thousands of dollars, or come from some poor country (like we do).

A highly efficient, real-time text to speech system deployed on CPUs

The Hateful Memes AI Challenge

We’ve built and are now sharing a data set designed specifically to help AI researchers develop new systems to identify multimodal hate speech. This content combines different modalities, such as text and images, making it difficult for machines to understand. The Hateful Memes data set contains 10,000+ new multimodal examples created by Facebook AI.

Ultimate Guide to Natural Language Processing Courses

Selecting an online course that will match your requirements is very frustrating if you have high standards. Most of them are not comprehensive and a lot of time spent on them is wasted. How would you feel, if someone would provide you a critical path and tell, what modules exactly and in which order will provide you comprehensive, expert-level knowledge?

Word2Vec: A Comparison Between CBOW, SkipGram & SkipGramSI

Learn how different Word2Vec architectures behave in practice. This is to help you make an informed decision on which architecture to use given the problem you are trying to solve. In this article, we will look at how the different neural network architectures for training a Word2Vec model behave in practice.

The Dark Secrets Of BERT

BERT stands for Bidirectional Encoder Representations from Transformers. This model is basically a multi-layer bidirectional Transformer encoder(Devlin, Chang, Lee, & Toutanova, 2019), and there are multiple excellent guides about how it works generally, includingthe Illustrated Transformer. What we focus on is one specific component of Transformer architecture known as self-attention.

10 Best Machine Learning Textbooks that All Data Scientists Should Read

The Best NLP Papers From ICLR 2020

I went through 687 papers that were accepted to ICLR 2020 virtual conference (out of 2594 submitted – up 63% since 2019!) and identified 9 papers with the potential to advance the use of deep learning NLP models in everyday use cases.

A Hacker’s Guide to Efficiently Train Deep Learning Models

Three months ago, I participated in a data science challenge that took place at my company. The goal was to help a marine researcher better identify whales based on the appearance of their flukes. More specifically, we were asked to predict for each image of a test set, the top 20 most similar images from the full database (train+test).

A state-of-the-art open source chatbot

Facebook AI has built and open-sourced Blender, the largest-ever open-domain chatbot. It outperforms others in terms of engagement and also feels more human, according to human evaluators. The culmination of years of research in conversational AI, this is the first chatbot to blend a diverse set of conversational skills — including empathy, knowledge, and personality — together in one system.

Google’s medical AI was super accurate in a lab. Real life was a different story.

Facebook AI, AWS partner to release new PyTorch libraries

Facebook AI and AWS have partnered to release new libraries that target high-performance PyTorch model deployment and large scale model training. As part of the broader PyTorch community, Facebook AI and AWS engineers have partnered to develop new libraries targeted at large-scale elastic and fault-tolerant model training and high-performance PyTorch model deployment. These libraries enable the community to efficiently productionize AI models at scale and push the state of the art on model exploration as model architectures continue to increase in size and complexity.

MIT CSAIL TextFooler Framework Tricks Leading NLP Systems

A team of researchers at the MIT Computer Science & Artificial Intelligence Lab (CSAIL) recently released a framework called TextFooler which successfully tricked state-of-the-art NLP models (such as BERT) into making incorrect predictions.

Google AI tool will no longer use gendered labels like ‘woman’ or ‘man’ in photos of people

When machine learning packs an economic punch

A new study co-authored by an MIT economist shows that improved translation software can significantly boost international trade online — a notable case of machine learning having a clear impact on economic activity. The research finds that after eBay improved its automatic translation program in 2014, commerce shot up by 10.9 percent among pairs of countries where people could use the new system. To put the results in perspective, he adds, consider that physical distance is, by itself, also a significant barrier to global commerce.

OpenAI, PyTorch

We are standardizing OpenAI’s deep learning framework on PyTorch. In the past, we implemented projects in many frameworks depending on their relative strengths. We’ve now chosen to standardize to make it easier for our team to create and share optimized implementations of our models.

Artificial intelligence yields new antibiotic

Using a machine-learning algorithm, MIT researchers have identified a powerful new antibiotic compound. In laboratory tests, the drug killed many of the world’s most problematic disease-causing bacteria, including some strains that are resistant to all known antibiotics. It also cleared infections in two different mouse models.

Deep Learning for Anomaly Detection

Anomalies, often referred to as outliers, are data points or patterns in data that do not conform to a notion of normal behavior. Anomaly detection, then, is the task of finding those patterns in data that do not adhere to expected norms. The capability to recognize or detect anomalous behavior can provide highly useful insights across industries.

2019 in Review: 10 AI Papers That Made an Impact

How we built the good first issues feature

We’ve recently launched good first issues recommendations to help new contributors find easy gateways into open source projects. Read about the machine learning engine behind these recommendations. GitHub is leveraging machine learning (ML) to help more people contribute to open source.

Introducing the AI Index 2019 Report

The AI Index 2019 Report takes an interdisciplinary approach by design, analyzing and distilling patterns about AI’s broad global impact on everything from national economies to job growth, research and public perception. We’re excited to release the AI Index 2019 Report, one of the most comprehensive studies about AI to date. Because AI touches so many aspects of society, the Index takes an interdisciplinary approach by design, analyzing and distilling patterns about AI’s broad global impact on everything from national economies to job growth, research and public perception.

Personalizing Spotify Home with Machine Learning

Machine learning is at the heart of everything we do at Spotify. Especially on Spotify Home, where it enables us to personalize the user experience and provide billions of fans the opportunity to enjoy and be inspired by the artists on our platform. This is what makes Spotify unique.

Building a document understanding pipeline with Google Cloud

Document understanding is the practice of using AI and machine learning to extract data and insights from text and paper sources such as emails, PDFs, scanned documents, and more. In the past, capturing this unstructured or “dark data” has been an expensive, time-consuming, and error-prone process requiring manual data entry. Today, AI and machine learning have made great advances towards automating this process, enabling businesses to derive insights from and take advantage of this data that had been previously untapped.

Google Research Use of Concept Vectors for Image Search

Google recently released research about creating a tool for searching Similar Medical Images Like Yours (SMILY). The research uses embeddings for image-based search and allows users to influence the search through the interactive refinement of concepts.

Introducing Hypothesis GU Funcs, an Open Source Python Package for Unit Testing

An introduction to audio processing and machine learning using Python

Top DevOps tools that you must use in 2019 and beyond

The Effects of Mixing Machine Learning and Human Judgment

In 1997 IBM’s Deep Blue software beat the World Chess Champion Garry Kasparov in a series of six matches. Since then, other programs have beaten human players in games ranging from Jeopardy to Go. Inspired by his loss, Kasparov decided in 2005 to test the success of Human+AI pairs in an online chess tournament.2

Understanding Convolutional Neural Networks

A Convolutional Neural Network (CNN) is a class of deep, feed-forward artificial neural networks most commonly applied to analyzing visual imagery. The architecture of these networks was loosely inspired by biological neurons that communicate with each other and generate outputs dependent on the inputs. Though work on CNNs started in the early 1980s, they only became popular with recent technology advancements and computational capabilities that allow the processing of large amounts of data and the training of sophisticated algorithms in a reasonable amount of time.

Three Approaches to Scaling Machine Learning with Uber Seattle Engineering

Introducing LCA: Loss Change Allocation for Neural Network Training

Neural networks (NNs) have become prolific over the last decade and now power machine learning across the industry. At Uber, we use NNs for a variety of purposes, including detecting and predicting object motion for self-driving vehicles, responding more quickly to customers, and building better maps. While many NNs perform quite well at their tasks, networks are fundamentally complex systems, and their training and operation is still poorly understood.

Replay in biological and artificial neural networks

Our waking and sleeping lives are punctuated by fragments of recalled memories: a sudden connection in the shower between seemingly disparate thoughts, or an ill-fated choice decades ago that haunts us as we struggle to fall asleep. By measuring memory retrieval directly in the brain, neuroscientists have noticed something remarkable: spontaneous recollections, measured directly in the brain, often occur as very fast sequences of multiple memories. These so-called ‘replay’ sequences play out in a fraction of a second–so fast that we’re not necessarily aware of the sequence.

Introducing KiloGram, a New Technique for AI Detection of Malware

Powered by AI: Oculus Insight

To unlock the full potential of virtual reality (VR) and augmented reality (AR) experiences, the technology needs to work anywhere, adapting to the spaces where people live and how they move within those real-world environments. When we developed Oculus Quest, the first all-in-one, completely wire-free VR gaming system, we knew we needed positional tracking that was precise, accurate, and available in real time — within the confines of a standalone headset, meaning it had to be compact and energy efficient. At last year’s Oculus Connect event we shared some details about Oculus Insight, the cutting-edge technology that powers both Quest and Rift S. Now that both of those products are available, we’re providing a deeper look at the AI systems and techniques that power this VR technology.

Deep probabilistic modelling with Pyro

Classical machine learning and deep learning algorithms can only propose the most probable solutions and are not able to adequately model uncertainty. The success of deep neural networks in diverse areas as image recognition and natural language processing has been outstanding in recent years. However, classical machine learning and deep learning algorithms can only propose the most probable solutions and are not able to adequately model uncertainty.

Speak to me: How voice commerce is revolutionizing commerce

We’ve seen profound advances in technology, especially with the development of artificial intelligence and deep learning which are increasingly for voice assistants. This, in turn, promises to bring about huge changes in consumer behavior — what’s being called “voice commerce”. This is a new channel, governed by a new set of rules.

New advances in natural language processing

Natural language understanding (NLU) and language translation are key to a range of important applications, including identifying and removing harmful content at scale and connecting people across different languages worldwide. Although deep learning–based methods have accelerated progress in language processing in recent years, current systems are still limited when it comes to tasks for which large volumes of labeled training data are not readily available. Recently, Facebook AI has achieved impressive breakthroughs in NLP using semi-supervised and self-supervised learning techniques, which leverage unlabeled data to improve performance beyond purely supervised systems.

Teaching Computers to Answer Complex Questions

Computerized question-answering systems usually take one of two approaches. Either they do a text search and try to infer the semantic relationships between entities named in the text, or they explore a hand-curated knowledge graph, a data structure that directly encodes relationships among entities. With complex questions, however — such as “Which Nolan films won an Oscar but missed a Golden Globe?” — both of these approaches run into difficulties.

First Programmable Memristor Computer

Michigan team builds memristors atop standard CMOS logic to demo a system that can do a variety of edge computing AI tasks Hoping to speed AI and neuromorphic computing and cut down on power consumption, startups, scientists, and established chip companies have all been looking to do more computing in memory rather than in a processor’s computing core. Memristors and other nonvolatile memory seem to lend themselves to the task particularly well. However, most demonstrations of in-memory computing have been in standalone accelerator chips that either are built for a particular type of AI problem or that need the off-chip resources of a separate processor in order to operate.

EGG: A toolkit for language emergence simulations with neural networks

Mapping roads through deep learning and weakly supervised training

Creating accurate maps today is a painstaking, time-consuming manual process, even with access to satellite imagery and mapping software. Many regions — particularly in the developing world — remain largely unmapped. To help close this gap, Facebook AI researchers and engineers have developed a new method that uses deep learning and weakly supervised training to predict road networks from commercially available high-resolution satellite imagery.

Introducing EvoGrad: A Lightweight Library for Gradient-Based Evolution

Tools that enable fast and flexible experimentation democratize and accelerate machine learning research. Take for example the development of libraries for automatic differentiation, such as Theano, Caffe, TensorFlow, and PyTorch: these libraries have been instrumental in catalyzing machine learning research, enabling gradient descent training without the tedious work of hand-computing derivatives. In these frameworks, it’s simple to experiment by adjusting the size and depth of a neural network, by changing the error function that is to be optimized, and even by inventing new architectural elements, like layers and activation functions–all without having to worry about how to derive the resulting gradient of improvement.

Panel: First Steps with Machine Learning

This panel is a very diverse group, and I’m actually going to let them introduce themselves rather than me trying to butcher any names. This is all about answering my need, literally, my first steps. What should I be focused on as a software engineer wanting to get into ML and start using ML more convinced leadership on things that I want to do?

Amenity Detection and Beyond—New Frontiers of Computer Vision at Airbnb

Grafana Labs Teams Use Jaeger to Improve Query Performance Up to 10x

Facebook AI Memory Layer Boosts Network Capacity by a Billion Parameters

Using natural language processing to manage healthcare records

The next time you see your physician, consider the times you fill in a paper form. It may seem trivial, but the information could be crucial to making a better diagnosis. Now consider the other forms of healthcare data that permeate your life—and that of your doctor, nurses, and the clinicians working to keep patients thriving.

How to run evolution strategies on Google Kubernetes Engine

Reinforcement learning (RL) has become popular in the machine learning community as more and more people have seen its amazing performance in games, chess and robotics. In previous blog posts we’ve shown you how to run RL algorithms on AI Platform utilizing both Google’s powerful computing infrastructure and intelligently managed training service such as Bayesian hyperparameter optimization. In this blog, we introduce Evolution Strategies (ES) and show how to run ES algorithms on Google Kubernetes Engine (GKE).

No Coding Required: Training Models with Ludwig, Uber’s Open Source Deep Learning Toolbox

AQR’s Problem With Machine Learning: Cats Morph Into Dogs

Machine learning has done magic, such as beating human chess champions. But in finance, expectations for the technology may need to come down a notch or two, according to quantitative firm AQR. Machine learning changes the way problems are solved.

How AI is Starting to Influence Wireless Communications

Machine learning and deep learning technologies are promising an end-to-end optimization of wireless networks while they commoditize PHY and signal-processing designs and help overcome RF complexities What happens when artificial intelligence (AI) technology arrives on wireless channels? For a start, AI promises to address the design complexity of radio frequency (RF) systems by employing powerful machine learning algorithms and significantly improving RF parameters such as channel bandwidth, antenna sensitivity and spectrum monitoring. So far, engineering efforts have been made for smartening individual components in wireless networks via technologies like cognitive radio.

Releasing Pythia for vision and language multimodal AI models

Pythia is a deep learning framework that supports multitasking in the vision and language domain. Built on our open-source PyTorch framework, the modular, plug-and-play design enables researchers to quickly build, reproduce, and benchmark AI models. Pythia is designed for vision and language tasks, such as answering questions related to visual data and automatically generating image captions.

Detecting malaria with deep learning

Artificial intelligence (AI) and open source tools, technologies, and frameworks are a powerful combination for improving society. ‘Health is wealth’ is perhaps a cliche, yet it’s very accurate! In this article, we will examine how AI can be leveraged for detecting the deadly disease malaria with a low-cost, effective, and accurate open source deep learning solution.

Untold History of AI: When Charles Babbage Played Chess With the Original Mechanical Turk

An ML showdown in search of the best tool

Ever burgeoning digital data combined with impressive research has lead to a rising interest in Machine Learning or ML, which has further powered a vibrant ecosystem of technologies, frameworks, and libraries in the space. Scikit-learn sees high adoption from the tech community. The most probable reason is a powerful Python interface that allows tweaking of models across multiple parameters.

Creating Bitcoin trading bots that don’t lose money

In this article we are going to create deep reinforcement learning agents that learn to make money trading Bitcoin. In this tutorial we will be using OpenAI’s gym and the PPO agent from the stable-baselines library, a fork of OpenAI’s baselines library. If you are not already familiar with how to create a gym environment from scratch, or how to render simple visualizations of those environments, I have just written articles on both of those topics.

DeepMind and Google: the battle to control artificial intelligence

One afternoon in August 2010, in a conference hall perched on the edge of San Francisco Bay, a 34-year-old Londoner called Demis Hassabis took to the stage. Walking to the podium with the deliberate gait of a man trying to control his nerves, he pursed his lips into a brief smile and began to speak: “So today I’m going to be talking about different approaches to building…” He stalled, as though just realising that he was stating his momentous ambition out loud.

Hash Your Way To a Better Neural Network

The computer industry has been busy in recent years trying to figure out how to speed up the calculations needed for artificial neural networks—either for their training or for what’s known as inference, when the network is performing its function. In particular, much effort has gone into designing special-purpose hardware to run such computations. Google, for example, developed its Tensor Processing Unit, or TPU, first described publicly in 2016.

12 open source tools for natural language processing

It would be easy to argue that Natural Language Toolkit (NLTK) is the most full-featured tool of the ones I surveyed. It implements pretty much any component of NLP you would need, like classification, tokenization, stemming, tagging, parsing, and semantic reasoning. And there’s often more than one implementation for each, so you can choose theexact algorithm or methodology you’d like to use.

How IBM Watson Overpromised and Underdelivered on AI Health Care

Using Machine Learning to Ensure the Capacity Safety of Individual Microservices

Introducing Ludwig, a Code-Free Deep Learning Toolbox

Over the last decade, deep learning models have proven highly effective at performing a wide variety of machine learning tasks in vision, speech, and language. At Uber we are using these models for a variety of tasks, including customer support, object detection, improving maps, streamlining chat communications, forecasting, and preventing fraud. Many open source libraries, including TensorFlow, PyTorch, CNTK, MXNET, and Chainer, among others, have implemented the building blocks needed to build such models, allowing for faster and less error-prone development.

Machine Learning-Powered Search Ranking of Airbnb Experiences

How we built and iterated on a machine learning Search Ranking platform for a new two-sided marketplace and how we helped itgrow. Airbnb Experiences are handcrafted activities designed and led by expert hosts that offer a unique taste of local scene and culture. Each experience is vetted for quality by a team of editors before it makes its way onto the platform.

Teaching AI to learn speech the way children do

A collaboration between the Facebook AI Research (FAIR) group and the Paris Sciences & Lettres University, with additional sponsorship from Microsoft Research, to challenge other researchers to teach AI systems to learn speech in a way that more closely resembles how young children learn. The ZeroSpeech 2019 challenge (which builds on previous efforts in 2015 and 2017) asks participants to build a speech synthesizer using only audio input, without any text or phonetic labels. The challenge’s central task is to build an AI system that can discover, in an unknown language, the machine equivalent of text of phonetic labels and use them to re-synthesize a sentence in a given voice.

Updating Neural Networks to Recognize New Categories, with Minimal Retraining

More-Efficient “Kernel Methods” Dramatically Reduce Training Time for Natural-Language-Understanding Systems

AI Blueprints: Implementing content-based recommendations using Python

In this article, we’ll have a look at how you can implement a content-based recommendation system using Python and the scikit-learn library. But before diving straight into this, it’s important to have some prerequisite knowledge of the different ways by which recommendation systems can recommend an item to users. Content-based: A content-based recommendation finds similar items to a given item by examining the item’s properties, such as its title or description, category, or dependencies on other items (for example, electronic toys require batteries).

Manifold: A Model-Agnostic Visual Debugging Tool for Machine Learning at Uber

Creating a Zoo of Atari-Playing Agents to Catalyze the Understanding of Deep Reinforcement Learning

AI year in review

At Facebook, we think that artificial intelligence that learns in new, more efficient ways – much like humans do – can play an important role in bringing people together. That core belief helps drive our AI strategy, focusing our investments in long-term research related to systems that learn using real-world data, inspiring our engineers to share cutting-edge tools and platforms with the wider AI community, and ultimately demonstrating new ways to use the technology to benefit the world. In 2018, we made important progress in all these areas.

What Kagglers are using for Text Classification

 With the problem of Image Classification is more or less solved by Deep learning, Text Classification is the next new developing theme in deep learning. For those who don’t know, Text classification is a common task in natural language processing, which transforms a sequence of text of indefinite length into a category of text. How could you use that?

Faster Neural Networks Straight from JPEG

Uber AI Labs introduces a method for making neural networks that process images faster and more accurately by leveraging JPEG representations. Neural networks, an important tool for processing data in a variety of industries, grew from an academic research area to a cornerstone of industry over the last few years. Convolutional Neural Networks (CNNs) have been particularly useful for extracting information from images, whether classifying them, recognizing faces, or evaluating board positions in Go.

DeepMind Achieves Holy Grail: An AI That Can Master Games Like Chess and Go Without Human Help

Using AI and satellite imagery for disaster insights

A framework for using convolutional neural networks (CNNs) on satellite imagery to identify the areas most severely affected by a disaster. This new method has the potential to produce more accurate information in far less time than current manual methods. Ultimately, the goal of this research is to allow rescue workers to quickly identify where aid is needed most, without relying on manually annotated, disaster-specific data sets.

Matplotlib—Making data visualization interesting

Data visualization is a key step to understand the dataset and draw inferences from it. While one can always closely inspect the data row by row, cell by cell, it’s often a tedious task and does not highlight the big picture. Visuals on the other hand, define data in a form that is easy to understand with just a glance and keeps the audience engaged.

Easy-To-Read Summary of Important AI Research Papers of 2018

Trying to keep up with AI research papers can feel like an exercise in futility given how quickly the industry moves. If you’re buried in papers to read that you haven’t quite gotten around to, you’re in luck. To help you catch up, we’ve summarized 10 important AI research papers from 2018 to give you a broad overview of machine learning advancements this year.

Amazon makes its machine learning courses available for free

Amazon announced today that it’s making its range of machine learning courses available to all developers signed up to its AWS platform for free. This program was previously available only to Amazon employees, but anyone can now take advantage of it at no charge by signing up to Amazon Web Services’ free plan. It includes 30 courses in total, with over 45 hours of course material, videos, and lab tests.

Montezuma’s Revenge Solved by Go-Explore, a New Algorithm for Hard-exploration Problems

FastMRI open source tools from Facebook and NYU

Facebook AI Research (FAIR) and NYU School of Medicine’s Center for Advanced Imaging Innovation and Research (CAI²R) are sharing new open source tools and data as part of fastMRI, a joint research project to spur development of AI systems to speed MRI scans by up to 10x. Today’s releases include new AI models and baselines for this task(as described in our paper here). It also includes the first large-scale MRI data set of its kind, which can serve as a benchmark for future research.

Humanizing Customer Complaints using NLP Algorithms

Last Christmas, I went through the most frustrating experience as a consumer. I was doing some last minute holiday shopping and after standing in a long line, I finally reached the blessed register only to find out that my debit card was blocked. I could sense the old lady at the register judging me with her narrowed eyes.

Radiology and Deep Learning

Radiology and DeepLearningDetecting pneumonia opacities from chest X-Ray images using deep learning. One day back in August, I was catching up with my best friend from high school who is now a radiology resident. One thing led to another, and we started talking about our interests in artificial intelligence and machine learning and its possible applications in radiology.

Predictive Scaling for EC2, Powered by Machine Learning

When I look back on the history of AWS and think about the launches that truly signify the fundamentally dynamic, on-demand nature of the cloud, two stand out in my memory: the launch of Amazon EC2 in 2006 and the concurrent launch of CloudWatch Metrics, Auto Scaling, and Elastic Load Balancing in 2009. The first launch provided access to compute power; the second made it possible to use that access to rapidly respond to changes in demand. We have added a multitude of features to all of these services since then, but as far as I am concerned they are still central and fundamental!

The dark side of YouTube

The YouTube algorithm that I helped build in 2011 still recommends the flat earth theory by the hundreds of millions. This investigation by @RawStory shows some of the real-life consequences of this badly designed AI.

The technology enabling Alibaba to sell $30.8 billion in Double 11 goods

Decision Tree in Machine Learning

A decision tree is a flowchart-like structure in which each internal node represents a test on a feature (e.g. whether a coin flip comes up heads or tails), each leaf node represents a class label (decision taken after computing all features) and branches represent conjunctions of features that lead to those class labels. The paths from root to leaf represent classification rules. Below diagram illustrate the basic flow of decision tree for decision making with labels (Rain(Yes), No Rain(No)).

Five Lessons From the First Three Years of Michelangelo

Uber has been one of the most active contributors to open source machine learning technologies in the last few years. While companies like Google or Facebook have focused their contributions in new deep learning stacks like TensorFlow, Caffe2 or PyTorch, the Uber engineering team has really focused on tools and best practices for building machine learning at scale in the real world. Technologies such as Michelangelo, Horovod, PyML, Pyro are some of examples of Uber’s contributions to the machine learning ecosystem.

Real Time Facial Expression Recognition

Computer animated agents and robots bring new dimension in human computer interaction which makes it vital as how computers can affect our social life in day-to-day activities. Face to face communication is a real-time process operating at a time scale in the order of milliseconds. The level of uncertainty at this time scale is considerable, making it necessary for humans and machines to rely on sensory rich perceptual primitives rather than slow symbolic inference processes.

Accurate Online Speaker Diarization with Supervised Learning

Speaker diarization, the process of partitioning an audio stream with multiple people into homogeneous segments associated with each individual, is an important part of speech recognition systems. By solving the problem of “who spoke when”, speaker diarization has applications in many important scenarios, such as understanding medical conversations, video captioning and more. However, training these systems with supervised learning methods is challenging — unlike standard supervised classification tasks, a robust diarization model requires the ability to associate new individuals with distinct speech segments that weren’t involved in training.

A Google Brain engineer’s guide to entering AI

Note that this guide was written in November 2018 to complement an in-depth conversation on the 80,000 Hours Podcast with Catherine Olsson and Daniel Ziegler on how to transition from computer science and software engineering in general into ML engineering, with a focus on alignment and safety. If you like this guide, we’d strongly encourage you to check out the podcast episode where we discuss some of the instructions here, and other relevant advice. Technical AI safety is a multifaceted area of research, with many sub-questions in areas such as reward learning, robustness, and interpretability.

How to Develop Convolutional Neural Network Models for Time Series Forecasting

Learning Concepts with Energy Functions

We’ve developed an energy-based model that can quickly learn to identify and generate instances of concepts, such as near, above, between, closest, and furthest, expressed as sets of 2d points. Our model learns these concepts after only five demonstrations.

20 Best YouTube channels for AI and machine learning

What are the most interesting and informative YouTube channels about artificial intelligence (AI) and machine learning? Subscribe to these 20 high-quality channels today to stay up to date with the latest AI and machine learning breakthroughs. Siraj Raval:

Why Chinese Artificial Intelligence Will Run The World

With Chinese tech giants Baidu, Alibaba, and Tencent focused on developing sophisticated AI-driven systems in the coming decade, the rest of the world can only watch while China builds the computer systems that will run our world in the decades to come. If you’ve been paying attention in the past year, it seems that all anyone can talk about is the coming artificial intelligence boom on the horizon. Whether it’s the Amazon, Google, or Facebook, everyone seems to be getting in on the AI game as fast as they can.

EPO Issues First Guidelines on AI Patents

The European Patent Office (EPO) has issued official guidelines on the patenting of artificial intelligence and machine learning technologies. The guidelines became valid on November 1st, 2018. When determining whether the claimed subject-matter satisfies this condition, the guidelines note that expressions such as “support vector machine,” “reasoning engine” or “neural network” may not qualify, as these are regarded as terms for mathematical methods which do not have a unique technical character of their own.

Tensorflow 2.0: models migration and new design

Tensorflow 2.0 will be a major milestone for the most popular machine learning framework: lots of changes are coming, and all with the aim of making ML accessible to everyone. These changes, however, requires for the old users to completely re-learn how to use the framework: this article describes all the (known) differences between the 1.x and 2.x version, focusing on the change of mindset required and highlighting the pros and cons of the new and implementations. This article can be a good starting point also for the novice: start thinking in the Tensorflow 2.0 way right now, so you don’t have to re-learn a new framework (unless until Tensorflow 3.0 will be released).

Horizon: An open-source reinforcement learning platform

Horizon is the first open source end-to-end platform that uses applied reinforcement learning (RL) to optimize systems in large-scale production environments. The workflows and algorithms included in this release were built on open frameworks — PyTorch 1.0, Caffe2, and Spark — making Horizon accessible to anyone using RL at scale. We’ve put Horizon to work internally over the past year in a wide range of applications, including helping to personalize M suggestions, delivering more meaningful notifications, and optimizing streaming video quality.

New Theory of Intelligence May Disrupt AI and Neuroscience

Recent advancement in artificial intelligence, namely in deep learning, has borrowed concepts from the human brain. The architecture of most deep learning models is based on layers of processing– an artificial neural network that is inspired by the neurons of the biological brain. Yet neuroscientists do not agree on exactly what intelligence is, and how it is formed in the human brain — it’s a phenomena that remains unexplained.

What’s the Best Deep Learning Framework?

Deep learning models are large and complex, so instead of writing out every function from the ground up, programmers rely on frameworks and software libraries to build neural networks efficiently. The top deep learning frameworks provide highly optimized, GPU-enabled code that are specific to deep neural network computations.

Curiosity and Procrastination in Reinforcement Learning

Episodic Curiosity through Reachability: Observations are added to memory, reward is computed based on how far the current observation is from the most similar observation in memory. The agent receives more reward for seeing observations which are not yet represented in memory.

Preview 7 Open Source Projects from the Uber Open Summit

Michelangelo PyML: Introducing Uber’s Platform for Rapid Python ML Model Development

California Law Bans Bots From Pretending to Be Human

Are you talking to a real person online or a bot? In California, bots will need to identify themselves thanks to a new bill just signed into law by Gov. Jerry Brown. The measure bans automated accounts from pretending to be real people in order to ‘incentivize a purchase or sale of goods or services in a commercial transaction or to influence a vote in an election,’ effective July 1, 2019.

DeepMind’s New Research on Linking Memories, and How It Applies to AI

Introducing Petastorm: Uber ATG’s Data Access Library for Deep Learning

AI Company Accused of Using Humans to Fake Its AI

Google AI Tool Identifies a Tumor’s Mutations From an Image

The What-If Tool: Code-Free Probing of Machine Learning Models

How would changes to a datapoint affect my model’s prediction? Does it perform differently for various groups–for example, historically marginalized people? How diverse is the dataset I am testing my model on?

Carnegie Mellon Researchers Develop New Deepfake Method

Deepfakes, ultrarealistic fake videos manipulated using machine learning, are getting pretty convincing. And researchers continue to develop new methods to create these types of videos, for better or, more likely, for worse. The most recent method comes from researchers at Carnegie Mellon University, who have figured out a way to automatically transfer the “style” of one person to another.

DARPA Starts $2 Billion Campaign to Develop Next Wave of AI Technologies

AI Hedge Fund Sentient Is Said to Shut After Less Than Two Years

AI System Approved For Diabetic Retinopathy Diagnosis

A system designed by a University of Iowa ophthalmologist that uses artificial intelligence (AI) to detect diabetic retinopathy without a person interpreting the results earned Food and Drug Administration (FDA) authorization in April, following a clinical trial in primary care offices. Results of that study were published Aug. 28 online in Nature Digital Medicine, offering the first look at data that led to FDA clearance for IDx-DR, the first medical device that uses AI for the autonomous detection of diabetic retinopathy. The clinical trial, which also was the first study to prospectively assess the safety of an autonomous AI system in patient care, compared the performance of IDx-DR to the gold standard diagnostic for diabetic retinopathy, which is the leading cause of vision loss in adults and one of the most severe complications for the 30.3 million Americans living with diabetes.

Deep Angel: AI that erases objects from images

Deep Angel is an artificial intelligence that erases objects from photographs. Part art, part technology, and part philosophy, Deep Angel shares Angelus Novus’ gaze into the future.

China Is Building a Fleet of Autonomous AI

A fleet of autonomous, AI-powered submarines is headed into hotly-contested Asian waterways. The vehicles will belong to the Chinese armed forces, and their mission capabilities are likely to raise concerned eyebrows in surrounding countries. If all goes to plan, the first submarines will launch in 2020.

Collection of Interactive Machine Learning Examples

Each seed is a machine learning example you can start playing with. Explore, learn and grow them into whatever you like.

Foundations Machine Learning

Bloomberg presents ‘Foundations of Machine Learning,’ a training course that was initially delivered internally to the company’s software engineers as part of its ‘Machine Learning EDU’ initiative. This course covers a wide variety of topics in machine learning and statistical modeling. The primary goal of the class is to help participants gain a deep understanding of the concepts, techniques and mathematical frameworks used by experts in machine learning.

Don’t Learn TensorFlow! Start with Keras or PyTorch Instead

So, you want to learn deep learning? Whether you want to start applying it to your business, base your next side project on it, or simply gain marketable skills – picking the right deep learning framework to learn is the essential first step towards reaching your goal. We strongly recommend that you pick either Keras or PyTorch.

Neural scene representation and rendering

There is more than meets the eye when it comes to how we understand a visual scene: our brains draw on prior knowledge to reason and to make inferences that go far beyond the patterns of light that hit our retinas. For example, when entering a room for the first time, you instantly recognise the items it contains and where they are positioned. If you see three legs of a table, you will infer that there is probably a fourth leg with the same shape and colour hidden from view.

Intel’s New Path to Quantum Computing

Intel’s director of quantum hardware, Jim Clarke, explains the company’s two quantum computing technologies The limits of Tangle Lake’s technology Silicon spin qubits and how far away they are The importance of cryogenic control electronics Top quantum computing applications What problems keeps him up at night AI vs. Quantum Computing: which will be more important? IEEE Spectrum: What’s special about Tangle Lake?

AI chip startup Wave to buy Silicon Valley old-timer MIPS

AI Nationalism

The last few years have seen developments in machine learning research and commercialisation that have been pretty astounding. As just a few examples: Image recognition starts to achieve human-level accuracy at complex tasks, for example skin cancer classification. Big steps forward in applying neural networks to machine translation at Baidu, Google, Microsoft etc.

The 50 Best Free Datasets for Machine Learning

What are some open datasets for machine learning? We at Gengo decided to create the ultimate cheat sheet for high quality datasets. These range from the vast (looking at you, Kaggle) or the highly specific (data for self-driving cars).

Improving Language Understanding with Unsupervised Learning

We’ve obtained state-of-the-art results on a suite of diverse language tasks with a scalable, task-agnostic system, which we’re also releasing. Our approach is a combination of two existing ideas: transformers and unsupervised pre-training. These results provide a convincing example that pairing supervised learning methods with unsupervised pre-training works very well; this is an idea that many have explored in the past, and we hope our result motivates further research into applying this idea on larger and more diverse datasets.

Attacks against machine learning – an overview

At a high level, attacks against classifiers can be broken down into three types: Adversarial inputs, which are specially crafted inputs that have been developed with the aim of being reliably misclassified in order to evade detection. Adversarial inputs include malicious documents designed to evade antivirus, and emails attempting to evade spam filters. Data poisoning attacks, which involve feeding training adversarial data to the classifier.

Why do neural networks generalize so poorly?

Deep convolutional network architectures are often assumed to guarantee generalization for small image translations and deformations. In this paper we show that modern CNNs (VGG16, ResNet50, and InceptionResNetV2) can drastically change their output when an image is translated in the image plane by a few pixels, and that this failure of generalization also happens with other realistic small image transformations. Furthermore, the deeper the network the more we see these failures to generalize.

Training a neural network in phase-change memory beats GPUs

Compared to a typical CPU, a brain is remarkably energy-efficient, in part because it combines memory, communications, and processing in a single execution unit, the neuron. A brain also has lots of them, which lets it handle lots of tasks in parallel. Attempts to run neural networks on traditional CPUs run up against these fundamental mismatches.

Learn Reinforcement Learning from scratch

Deep RL is a field that has seen vast amounts of research interest, including learning to play Atari games, beating pro players at Dota 2, and defeating Go champions. Contrary to many classical Deep Learning problems that often focus on perception (does this image contain a stop sign?) , Deep RL adds the dimension of actions that influence the environment (what is the goal, and how do I get there?).

Horovod: Distributed Training Framework for TensorFlow, Keras, and PyTorch

Horovod is a distributed training framework for TensorFlow, Keras, and PyTorch. The goal of Horovod is to make distributed Deep Learning fast and easy to use.

AI winter is well on its way

Deep learning has been at the forefront of the so called AI revolution for quite a few years now, and many people had believed that it is the silver bullet that will take us to the world of wonders of technological singularity (general AI). Many bets were made in 2014, 2015 and 2016 when still new boundaries were pushed, such as the Alpha Go etc. Companies such as Tesla were announcing through the mouths of their CEO’s that fully self driving car was very close, to the point that Tesla even started selling that option to customers [to be enabled by future software update].

Americans Less Trusting of Self-Driving Safety Following High-Profile Accidents

Americans are less trusting of self-driving cars following two deadly accidents involving autonomous or semi-autonomous vehicles, with half of U.S. adults considering those automobiles less safe than human drivers, according to a new poll. A Morning Consult survey conducted March 29-April 1 among a national sample of 2,202 adults found that 27 percent of respondents said self-driving cars are safer than human drivers, while 50 percent said autonomous vehicles are less safe. Eight percent said the automobiles are on par with human drivers when it comes to safety.

Making driverless cars change lanes more like human drivers do

The US government wants to start charging for the best free satellite data on earth

OpenAI: Gym Retro

We’re releasing the full version of Gym Retro, a platform for reinforcement learning research on games. This brings our publicly-released game count from around 70 Atari games and 30 Sega games to over 1,000 games across a variety of backing emulators. We’re also releasing the tool we use to add new games to the platform.

30+ Machine Learning Resources

For almost all machine learning projects, the main steps of the ideal solution remain same. Briefly, we all go over the steps below each and every time: Understand the dataClean up, fix the missing values, extract new features, select the best onesBuild the model, compare it with the other ones, tune hyper parameters, find out what is the right metric to evaluate your modelIterate this process over and over again until you believe you have the best solution:) Iterate this process over and over again until you believe you have the best solution:)

Intel AI Lab open-sources library for deep learning-driven NLP

The Intel AI Lab has open-sourced a library for natural language processing to help researchers and developers give conversational agents like chatbots and virtual assistants the smarts necessary to function, such as name entity recognition, intent extraction, and semantic parsing to identify the action a person wants to take from their words. The first-ever conference by Intel for AI developers is being held Wednesday and Thursday, May 23 and 24, at the Palace of Fine Arts in San Francisco. The Intel AI Lab now employs about 40 data scientists and researchers and works with divisions of the company developing products like the nGraph framework and hardware like Nervana Neural Network chips, Liu said.

3D Face Reconstruction with Position Map Regression Networks

Position Map Regression Networks (PRN) is a method to jointly regress dense alignment and 3D face shape in an end-to-end manner. In this article, I’ll provide a short explanation and discuss its applications in computer vision. In the last few decades, a lot of important research groups in computer vision have made amazing advances in 3D face reconstruction and face alignment.

Deep Learning Research: Creating Adaptable Meta-Learning Models

Adaptability is one of the key cognitive abilities that defined us as humans. Even as babies, we can intuitively shift between similar tasks even if we don’t have prior training on them. This contrasts with the traditional train-and-test approach of most artificial intelligence(AI) systems which require an agent to go through massive amounts of training before it can master a specific task.

Scanner: Processing Terabytes of Video on Hundreds of Machines

There are now many state-of-the-art computer vision algorithms which are a git clone away. We found that existing systems for distributed data analysis were not well suited to dealing with the computational challenges of applying computer vision algorithms to terabyte or petabyte sized video collections, so we designed and built a system called Scanner to make efficient video analysis easier.

Tensor Compilers: Comparing PlaidML, Tensor Comprehensions, and TVM

One of the most complex and performance critical parts of any machine learning framework is its support for device specific acceleration. Indeed, without efficient GPU acceleration, much of modern ML research and deployment would not be possible. This acceleration support is also a critical bottleneck, both in terms of adding support for a wider range of hardware targets (including mobile) as well as for writing new research kernels.

A Deep Dive into Monte Carlo Tree Search

The very first Go AIs used multiple modules to handle each aspect of playing Go – life and death, capturing races, opening theory, endgame theory, and so on. The idea was that by having experts program each module using heuristics, the AI would become an expert in all areas of the game. All that came to a grinding halt with the introduction of Monte Carlo Tree Search (MCTS) around 2008.

To Build Truly Intelligent Machines, Teach Them Cause and Effect

What the History of Math Can Teach Us About the Future of AI

Crossbar Pushes Resistive RAM into Embedded AI

Resistive RAM technology developer Crossbar says it has inked a deal with aerospace chip maker Microsemi allowing the latter to embed Crossbar’s nonvolatile memory on future chips. The move follows selection of Crossbar’s technology by a leading foundry for advanced manufacturing nodes. Crossbar is counting on resistive RAM (ReRAM) to enable artificial intelligence systems whose neural networks are housed within the device rather than in the cloud.

AI and Compute

We’re releasing an analysis showing that since 2012, the amount of compute used in the largest AI training runs has been increasing exponentially with a 3.5 month-doubling time (by comparison, Moore’s Law had an 18-month doubling period). Since 2012, this metric has grown by more than 300,000x (an 18-month doubling period would yield only a 12x increase). Improvements in compute have been a key component of AI progress, so as long as this trend continues, it’s worth preparing for the implications of systems far outside today’s capabilities.

Finland offers free online Artificial Intelligence course to anyone, anywhere

The Nengo Neural Simulator

Nengo is a graphical and scripting based Python package for simulating large-scale neural networks. Nengo can create sophisticated spiking or non-spiking neural simulations with sensible defaults in a few lines of code. Yet, Nengo is highly extensible and flexible.

Introducing state of the art text classification with universal language models

How an AI Startup Could Defeat Now Unbeatable Bugs

The need for new medications is higher than ever, but so is the cost and time to bring them to market. Developing a new drug can cost billions and take as long as 14 years, according to the U.S. Food and Drug Administration. Yet with all that effort, only 8 percent of drugs make it to market, the FDA said.

Prefrontal cortex as a meta-reinforcement learning system

Recently, AI systems have mastered a range of video-games such as Atari classics Breakout and Pong. But as impressive as this performance is, AI still relies on the equivalent of thousands of hours of gameplay to reach and surpass the performance of human video game players. In contrast, we can usually grasp the basics of a video game we have never played before in a matter of minutes.

Transfer Learning

Transfer Learning is the reuse of a pre-trained model on a new problem. It is currently very popular in the field of Deep Learning because it enables you to train Deep Neural Networks with comparatively little data. This is very useful since most real-world problems typically do not have millions of labeled data points to train such complex models.

An Introduction to Hashing in the Era of Machine Learning

What’s New in Deep Learning Research: Inside Google’s Semantic Experiences

Machine Learning for Text Classification Using SpaCy in Python

spaCy is a popular and easy-to-use natural language processing library in Python. It provides current state-of-the-art accuracy and speed levels, and has an active open source community. However, since SpaCy is a relative new NLP library, and it’s not as widely adopted as NLTK.

The fall of RNN / LSTM

It is the year 2014 and LSTM and RNN make a great come-back from the dead. We all read Colah’s blog and Karpathy’s ode to RNN. But we were all young and unexperienced.

Germany adopts first ethics standards for autonomous driving systems

Federal transport minister, Alexander Dobrindt, presented a report to Germany’s cabinet seeking to establish guidelines for the future programming of ethical standards into automated driving software. The report, was prepared by an automated driving ethics commission comprised of scientists and legal experts and produced 20 guidelines to be used by the automotive industry when creating automated driving systems. Shortly after its introduction, Dobrindt announced that the cabinet ratified the guidelines, making Germany the first government in the world to put such measures in place.

Automatic Photography with Google Clips

How could we train an algorithm to recognize interesting moments? As with most machine learning problems, we started with a dataset. We created a dataset of thousands of videos in diverse scenarios where we imagined Clips being used.

Custom deep learning loss functions with Keras for R

I recently started reading “Deep Learning with R”, and I’ve been really impressed with the support that R has for digging into deep learning. One of the use cases presented in the book is predicting prices for homes in Boston, which is an interesting problem because homes can have such wide variations in values. This is a machine learning problem that is probably best suited for classical approaches, such as XGBoost, because the data set is structured rather than perceptual data.

AI Can Generate ‘Doom’ Levels Now

Researchers recently successfully trained neural networks to generate level maps for Doom that, they report in a paper published to the arXiv preprint server in April, “proved to be interesting” to play. The work was carried out by researchers from the Polytechnic University of Milan and used Generative Adversarial Networks, a recent innovation in the field of deep learning.

AI trained to navigate develops brain-like location tracking

Now that DeepMind has solved Go, the company is applying DeepMind to navigation. Navigation relies on knowing where you are in space relative to your surroundings and continually updating that knowledge as you move. DeepMind scientists trained neural networks to navigate like this in a square arena, mimicking the paths that foraging rats took as they explored the space.

Intel Starts R&D Effort in Probabilistic Computing for AI

Intel announced today that it is forming a strategic research alliance to take artificial intelligence to the next level. Autonomous systems don’t have good enough ways to respond to the uncertainties of the real world, and they don’t have a good enough way to understand how the uncertainties of their sensors should factor into the decisions they need to make. According to Intel CTO Mike Mayberry the answer is “probabilistic computing”, which he says could be AI’s next wave.

Google’s new conversational AI could eventually undermine our sense of identity

Artificial Neural Nets Grow Brainlike Navigation Cells

Having the sense to take a shortcut, the most direct route from point A to point B, doesn’t sound like a very impressive test of intelligence. Yet according to a new report appearing today in Nature, in which researchers describe the performance of their new navigational artificial intelligence, the system’s ability to explore complex simulated environments and find the shortest route to a goal put it in a class previously reserved for humans and other living things. The surprising key to the system’s performance was that while learning how to navigate, the neural net spontaneously developed the equivalent of “grid cells,” sets of brain cells that enable at least some mammals to track their location in space.

Google launches the third version of its A.I. chips, an alternative to Nvidia’s

Google Duplex will call salons, restaurants, and pretend to be human for you [Updated]

Delivery Robots Will Rely on Human Kindness and Labor

In April, Starship Technologies announced that it is going to launch “robot delivery services for campuses.” Its goal is to deploy at least 1,000 delivery robots to “corporate and academic campuses in Europe and the U.S.” within the next year. It’s the latest in a long list of automated delivery schemes from a tech companies big and small.

Google Duplex: An AI System for Accomplishing Real World Tasks Over the Phone

Google: Deep Learning for Electronic Health Records

When patients get admitted to a hospital, they have many questions about what will happen next. When will I be able to go home? Will I get better?

Facebook’s Field Guide to Machine Learning video series

The Facebook Field Guide to Machine Learning is a six-part video series developed by the Facebook ads machine learning team. The series shares best real-world practices and provides practical tips about how to apply machine-learning capabilities to real-world problems. Machine learning and artificial intelligence are in the headlines everywhere today, and there are many resources to teach you about how the algorithms work and demonstrations of the latest cutting-edge research.

Cutting Edge Deep Learning for Coders, Part 2

Welcome to the new 2018 edition of fast.ai’s second 7 week course, Cutting Edge Deep Learning For Coders, Part 2, where you’ll learn the latest developments in deep learning, how to read and implement new academic papers, and how to solve challenging end-to-end problems such as natural language translation. You’ll develop a deep understanding of neural network foundations, the most important recent advances in the fields, and how to implement them in the world’s fastest deep learning libraries, fastai and pytorch. This course contains all new material, so if you’ve already completed the 2017 version, you’ll find plenty here to keep you busy too!

MIT Invented a Tool That Allows Driverless Cars to Navigate Rural Roads Without a Map

Real-Time AI: Microsoft Announces Preview of Project Brainwave

That’s where Microsoft’s Project Brainwave could come in. Project Brainwave is a hardware architecture designed to accelerate real-time AI calculations. The Project Brainwave architecture is deployed on a type of computer chip from Intel called a field programmable gate array, or FPGA, to make real-time AI calculations at competitive cost and with the industry’s lowest latency, or lag time.

Facial recognition tech wrongly identified 2,000 people as potential criminals

The economics of artificial intelligence

When looking at artificial intelligence from the perspective of economics, we ask the same, single question that we ask with any technology: What does it reduce the cost of? Economists are good at taking the fun and wizardry out of technology and leaving us with this dry but illuminating question. The answer reveals why AI is so important relative to many other exciting technologies.

16-year-old on finding primes with neural networks

Inspired by the triumphs of the “AlphaGo” project by DeepMind, I focused my research into the extensions and optimisation techniques that are so common in neural network design. There are many useful packages in machine learning, such as tensor flow, which can generate complex neural networks which work well, very quickly, but for this project I really wanted to develop an understanding of the inner workings of modern neural networks. So I went through the arduous calculus myself giving a much less efficient but more rewarding program.

Having AI systems try to outwit one another could help judge their intentions

Announcing PyTorch 1.0 for both research and production

PyTorch 1.0 takes the modular, production-oriented capabilities from Caffe2 and ONNX and combines them with PyTorch’s existing flexible, research-focused design to provide a fast, seamless path from research prototyping to production deployment for a broad range of AI projects. With PyTorch 1.0, AI developers can both experiment rapidly and optimize performance through a hybrid front end that seamlessly transitions between imperative and declarative execution modes. The technology in PyTorch 1.0 has already powered many Facebook products and services at scale, including performing 6 billion text translations per day.

Advancing state-of-the-art image recognition with deep learning on hashtags

Image recognition is one of the pillars of AI research and an area of focus for Facebook. Our researchers and engineers aim to push the boundaries of computer vision and then apply that work to benefit people in the real world — for example, using AI to generate audio captions of photos for visually impaired users. In order to improve these computer vision systems and train them to consistently recognize and classify a wide range of objects, we need data sets with billions of images instead of just millions, as is common today.

Embodied Question Answering: A goal-driven approach to autonomous agents

Facebook AI Research (FAIR) has developed a collection of virtual environments for training and testing autonomous agents, as well as novel AI agents that learn to intelligently explore those environments. To test this goal-driven approach, FAIR are collaborating Georgia Tech on a multistep AI task called Embodied Question Answering, or EmbodiedQA.

ONNX expansion speeds AI development

Facebook helped develop the Open Neural Network Exchange (ONNX) format to allow AI engineers to more easily move models between frameworks without having to do resource-intensive custom engineering. Today, we’re sharing that ONNX is adding support for additional AI tools, including Apple Core ML converter technology, Baidu’s PaddlePaddle platform, and Qualcomm SNPE.

We Need Bug Bounties for Bad Algorithms

Algorithmic auditors are a growing discipline of researchers specializing in computer science and human-computer interaction. They employ a variety of methods to tinker with and uncover how algorithms work, and their research has already sparked public discussions and regulatory investigations into the most dominant and powerful algorithms of the Information Age. From Uber and Booking.com to Google and Facebook, to name a few, these friendly auditors already uncovered bias and deception in the algorithms that control our lives.

The Army Is Working on Brain Hacks to Help Soldiers Deal With Information Overload

Facebook Open Sources ELF OpenGo

Inspired by DeepMind’s work, we kicked off an effort earlier this year to reproduce their recent AlphaGoZero results using FAIR’s Extensible, Lightweight Framework (ELF) for reinforcement learning research. The goal was to create an open source implementation of a system that would teach itself how to play Go at the level of a professional human player or better. By releasing our code and models we hoped to inspire others to think about new applications and research directions for this technology.

Artificial Intelligence Opens the Vatican Secret Archives

Known as In Codice Ratio, it uses a combination of artificial intelligence and optical-character-recognition (OCR) software to scour these neglected texts and make their transcripts available for the very first time. If successful, the technology could also open up untold numbers of other documents at historical archives around the world.

What tech calls “AI” isn’t really AI

First, the problem itself is poorly defined: what do you mean by intelligence? Nature, with all her blind hideous strength, endless experimentation and wild wastes of infinite time, has only managed the trick once (by our narrow definition), with one species of tree-ape on a rolling green world. Even if you believe there’s intelligent biological life elsewhere, the stats aren’t promising.

The EU is trying to decide whether to grant robots personhood.

Measuring the Intrinsic Dimension of Objective Landscapes

In our paper, Measuring the Intrinsic Dimension of Objective Landscapes, to be presented at ICLR 2018, we contribute to this ongoing effort by developing a simple way of measuring a fundamental network property known as intrinsic dimension. In the paper, we develop intrinsic dimension as a quantification of the complexity of a model in a manner decoupled from its raw parameter count, and we provide a simple way of measuring this dimension using random projections. We find that many problems have smaller intrinsic dimension than one might suspect.

DeepMind papers at ICLR 2018

Between 30 April and 03 May, hundreds of researchers will gather in Vancouver, Canada, for the Sixth International Conference on Learning Representations. Here you will find details of all DeepMind’s accepted papers.

Forging Voices and Faces: The Dangers of Audio and Video Fabrication

Pentagon-funded research aims to predict when crimes are gang-related

The paper attempts to predict whether crimes are gang-related using a neural network, a complex computational system modeled after a human brain that “learns” to classify or identify items based on ingesting a training dataset. The authors selected what they determined to be the four most important features (number of suspects, primary weapon used, the type of premises where the crime took place, and the narrative description of the crime) for identifying a gang-related crime from 2014–16 LAPD data and cross-referenced the crime incidents with a 2009 LAPD map of gang territory to create a training dataset for their neural network.

TDM: From Model-Free to Model-Based Deep Reinforcement Learning

While simple, this thought experiment highlights some important aspects of human intelligence. For some tasks, we use a trial-and-error approach, and for others we use a planning approach. A similar phenomena seems to have emerged in reinforcement learning (RL).

Swift for TensorFlow Design Overview

Swift for TensorFlow provides a new programming model for TensorFlow – one that combines the performance of graphs with the flexibility and expressivity of Eager execution, while keeping a strong focus on improved usability at every level of the stack. In order to achieve our goals, we scope in the possibility of making (carefully considered) compiler and language enhancements, which enables us to provide a great user experience.

Introduction to Decision Tree Learning

From Kaggle to classrooms, one of the first lessons in machine learning involves decision trees. The reason for the focus on decision trees is that they aren’t very mathematics heavy compared to other ML approaches, and at the same time, they provide reasonable accuracy on classification problems.

New A.I. application can write its own code

Designing applications that can program computers is a long-sought grail of the branch of computer science called artificial intelligence (AI). The new application, called Bayou, came out of an initiative aimed at extracting knowledge from online source code repositories like GitHub. Users can try it out at askbayou.com.

An open source implementation of DeepVoice 3: 2000-Speaker Neural Text-to-Speech

Why data scientists should start learning Swift

One week into my first year physics course at the University of Michigan, a professor assigned a problem set that required simulating some many-body system. It was due Friday. That was the week I learned my first programming language, Matlab.

United Kingdom Plans $1.3 Billion Artificial Intelligence Push

The U.K. government said Thursday that part of its multi-year AI investment–about £300 million, or more than $400 million–would come from U.K.-based corporations and investment firms and those located outside the country.

Aiming to fill skill gaps in AI, Microsoft makes training courses available to the public

Lessons from My First Two Years of AI Research

A friend of mine who is about to start a career in artificial intelligence research recently asked what I wish I had known when I started two years ago. Below are some lessons I have learned so far. They range from general life lessons to relatively specific tricks of the AI trade.

CIA plans to replace spies with AI

Human spies will soon be relics of the past, and the CIA knows it. Dawn Meyerriecks, the Agency’s deputy director for technology development, recently told an audience at an intelligence conference in Florida the CIA was adapting to a new landscape where its primary adversary is a machine, not a foreign agent.

Scientists plan huge European AI hub to compete with US

An Introduction to Hashing in the Era of Machine Learning

New AI Imaging Technique Reconstructs Photos with Realistic Results

To prepare to train their neural network, the team first generated 55,116 masks of random streaks and holes of arbitrary shapes and sizes for training. They also generated nearly 25,000 for testing. These were further categorized into six categories based on sizes relative to the input image, in order to improve reconstruction accuracy.

Accelerating Deep Neuroevolution: Train Atari in Hours on a Single Personal Computer

A new way to automatically build road maps from aerial images

A Face-Detection Library in 200 Lines of JavaScript

The pico.js library is a JavaScript implementation of the method described in 2013 by Markuš et al. in a technical report. The reference implementation is written in C and available on GitHub: https://github.com/nenadmarkus/pico.

The AI Revolution Hasn’t Happened Yet

Artificial Intelligence (AI) is the mantra of the current era. The phrase is intoned by technologists, academicians, journalists and venture capitalists alike. As with many phrases that cross over from technical academic fields into general circulation, there is significant misunderstanding accompanying the use of the phrase.

Machine Learning’s ‘Amazing’ Ability to Predict Chaos

Half a century ago, the pioneers of chaos theory discovered that the “butterfly effect” makes long-term prediction impossible. Even the smallest perturbation to a complex system (like the weather, the economy or just about anything else) can touch off a concatenation of events that leads to a dramatically divergent future. Unable to pin down the state of these systems precisely enough to predict how they’ll play out, we live under a veil of uncertainty.

China’s Strategic Ambiguity and Shifting Approach to Lethal Autonomous Weapons Systems

A step-by-step guide to the “World Models” AI paper

Teaching a machine to master car racing and fireball avoidance through “World Models”.

IBM Releases Open Source AI Security Tool

IBM releases Adversarial Robustness Toolbox, an open source software library designed to help researchers and developers secure artificial intelligence (AI) systems

China’s tech giants are venturing into autonomous driving

If there’s one thing China’s tech giants are known for, it’s their ability to venture into everything from social media, to online payments, to delivery services. The latest thing they’re all speeding towards? Autonomous driving.

This AI Will Turn Your Dog Into a Cat

As detailed in a paper published to arXiv, the neural net is actually a generative adversarial network (GAN), which is a way of training a machine learning algorithm without human supervision. In GANs, two neural nets are pitted against one another: One neural net generates new images and tries to trick the other neural net into thinking the images are real. If the other neural net is able to tell the generated images are false

Watch artificial intelligence create a 3D model of a person—from just a few seconds of video

Robot cognition requires machines that both think and feel

In the quest to create intelligent robots, designers tend to focus on purely rational, cognitive capacities. It’s tempting to disregard emotion entirely, or include only as much as necessary. But without emotion to help determine the personal significance of objects and actions, I doubt that true intelligence can exist – not the kind that beats human opponents at chess or the game of Go, but the sort of smarts that we humans recognise as such.

Google works out a fascinating, slightly scary way for AI to isolate voices in a crowd

15 Types of Regression you should know

Regression techniques are one of the most popular statistical techniques used for predictive modeling and data mining tasks. On average, analytics professionals know only 2-3 types of regression which are commonly used in real world. They are linear and logistic regression.

Someone has entered an AI in a Japanese mayoral race

How Lyft, Mastercard, and Drone Companies Are Experimenting With Artificial Intelligence

FDA permits marketing of artificial intelligence-based device to detect certain diabetes

Depthwise separable convolutions for machine learning

Convolutions are an important tool in modern deep neural networks (DNNs). This post is going to discuss some common types of convolutions, specifically regular and depthwise separable convolutions. My focus will be on the implementation of these operation, showing from-scratch Numpy-based code to compute them and diagrams that explain how things work.

Looking to Listen: Audio-Visual Speech Separation

People are remarkably good at focusing their attention on a particular person in a noisy environment, mentally “muting” all other voices and sounds. Known as the cocktail party effect, this capability comes natural to us humans. However, automatic speech separation — separating an audio signal into its individual speech sources — while a well-studied problem, remains a significant challenge for computers.

Gift of Garb: How AI Helps Fashion Followers Choose the Best Dress

If AI Thinks Like a Human It May Get Depressed

Towards a Virtual Stuntman

Motion control problems have become standard benchmarks for reinforcement learning, and deep RL methods have been shown to be effective for a diverse suite of tasks ranging from manipulation to locomotion. However, characters trained with deep RL often exhibit unnatural behaviours, bearing artifacts such as jittering, asymmetric gaits, and excessive movement of limbs. Can we train our characters to produce more natural behaviours?

This Chinese Facial Recognition Surveillance Company Is Now the World’s Most Valuable AI Startup

Differentiable Plasticity: A New Method Learning to Learn

Neural networks, which underlie many of Uber’s machine learning systems, have proven highly successful in solving complex problems, including image recognition, language understanding, and game-playing. However, these networks are usually trained to a stopping point through gradient descent, which incrementally adjusts the connections of the network based on its performance over many trials. Once the training is complete, the network is fixed and the connections can no longer change; as a result, barring any later re-training (again requiring many examples), the network in effect stops learning at the moment training ends.

The Case Against an Autonomous Military

The potential harm of A.I.s deliberately designed to kill in warfare is much more pressing. The U.S. and other countries are working hard to develop military A.I., in the form of automated weapons, that enhance battlefield capabilities while exposing fewer soldiers to injury or death. For the U.S., this would be a natural extension of the existing imperfect drone warfare program—failures in military intelligence have led to the mistaken killing of non-combatants in Iraq.

DeepMarks: A Digital Fingerprinting Framework for Deep Neural Networks

DeepMarks introduces the first fingerprinting methodology that enables the model owner to embed unique fingerprints within the parameters (weights) of her model and later identify undesired usages of her distributed models. The proposed framework embeds the fingerprints in the Probability Density Function (pdf) of trainable weights by leveraging the extra capacity available in contemporary DL models. DeepMarks is robust against fingerprints collusion as well as network transformation attacks, including model compression and model fine-tuning.

Lessons Learned Reproducing a Deep Reinforcement Learning Paper

There are a lot of neat things going on in deep reinforcement learning. One of the coolest things from last year was OpenAI and DeepMind’s work on training an agent using feedback from a human rather than a classical reward signal. There’s a great blog post about it at Learning from Human Preferences, and the original paper is at Deep Reinforcement Learning from Human Preferences.

Spyce Kitchen, a robotic chef built by four recent M.I.T. grads

AlterEgo: Interfacing with devices through silent speech

AlterEgo is a closed-loop, non-invasive, wearable system that allows humans to converse in high-bandwidth natural language with machines, artificial intelligence assistants, services, and other people without any voice—without opening their mouth, and without any discernible movements—simply by vocalizing internally.

More Power, Less Tower: AI May Make Aircraft Control Towers Obsolete

Berkeley offers its fastest-growing course – data science – online for free

AI Experts Threaten to Boycott a University Over Reports of Killer Robot Research

Retro Contest

We’re launching a transfer learning contest that measures a reinforcement learning algorithm’s ability to generalize from previous experience. In typical RL research, algorithms are tested in the same environment where they were trained, which favors algorithms which are good at memorization and have many hyperparameters. Instead, our contest tests an

Over 3,000 Google Employees Signed a Letter Demanding Google Leave the ‘Business of War’

Will This “Neural Lace” Brain Implant Help Us Compete with AI?

CS224n: Natural Language Processing with Deep Learning

NLP projects at Stanford.

Apple has hired Google’s head of search and artificial intelligence

Book: Mathematics for Machine Learning

We are in the process of writing a book on Mathematics for Machine Learning that motivates people to learn mathematical concepts. The book is not intended to cover advanced machine learning techniques because there are already plenty of books doing this. Instead, we aim to provide the necessary mathematical skills to read those other books.

Microsoft AI Interview Questions – Acing the AI Interview

In a recent letter by the CEO Satya Nadella to all Microsoft employees, there are two new teams formed in Microsoft, Intelligent Cloud and Intelligent Edge to shape the next phase of innovation. This announces the profound shift to weave Artificial Intelligence in to all that Microsoft does. Needless to say Microsoft following this announcement may increase the AI related hires to the company.

Machine Behavior Needs to Be an Academic Discipline

What if physiologists were the only people who study human behavior at all scales: from how the human body functions, to how social norms emerge, to how the stock market functions, to how we create, share, and consume culture? What if neuroscientists were the only people tasked with studying criminal behavior, designing educational curricula, and devising policies to fight tax evasion?

Norman: Psychopath AI

We present you Norman, world’s first psychopath AI. Norman suffered from extended exposure to the darkest corners of Reddit, and represents a case study on the dangers of Artificial Intelligence gone wrong when biased data is used in machine learning algorithms.

Turn TensorFlow functions into mathematical notations and diagrams

This is based on some helper classes I started writing, to help my self make less mistakes and understand the code better. It is still work in progress. I want to share the idea to see if it interests others.

Automatically “block” people in images using a pretrained neural network

Swift for TensorFlow

Swift for TensorFlow is a result of first-principles thinking applied to machine learning frameworks, and works quite differently than existing TensorFlow language bindings. Whereas prior solutions are designed within the constraints of what can be achieved by a (typically Python or Lua) library, Swift for TensorFlow is based on the belief that machine learning is important enough to deserve first-class language and compiler support.

Google’s TensorFlow AI framework adds Swift and JavaScript support

Google todayunveiled a slew of updates to its popular TensorFlow machine learning framework aimed at making it useful for a wider variety of developers and providing data scientists with new ways to get started building AI models.

French government’s new AI strategy gets a boost from DeepMind, Samsung, and Fujitsu

Project Clara: NVIDIA Supercomputing Platform Redefines Medical Imaging

NVIDIA’s Project Clara, a medical imaging supercomputer, renews the capabilities of these machines in place. Unveiled this week at the GPU Technology Conference, in Silicon Valley, Project Clara takes advantage of incredible advancements in computation.

Jensen Huang on the Uber Tragedy and Why Nvidia Suspended Testing

Introducing TensorFlow.js: Machine Learning in JavaScript

We’re excited to introduce TensorFlow.js, an open-source library you can use to define, train, and run machine learning models entirely in the browser, using Javascript and a high-level layers API. If you’re a Javascript developer who’s new to ML, TensorFlow.js is a great way to begin learning. Or, if you’re a ML developer who’s new to Javascript, read on to learn more about new opportunities for in-browser ML.

Watch a Human Mind-Control a Cyborg Mouse

Researchers from the Korea Advanced Institute of Science and Technology created steerable cyborg mice, and holy hell is this experiment giving me the creeps.

Interpretable Machine Learning: A Guide for Making Black Box Models Explainable

Universal Sentence Encoder by Ray Kurzweil’s Team at Google

We present models for encoding sentences into embedding vectors that specifically target transfer learning to other NLP tasks. The models are efficient and result in accurate performance on diverse transfer tasks. Two variants of the encoding models allow for trade-offs between accuracy and compute resources.

Teaching machines to spot essential information in physical systems

Two physicists at ETH Zurich and the Hebrew University of Jerusalem have developed a novel machine-learning algorithm that analyses large data sets describing a physical system and extract from them the essential information needed to understand the underlying physics.

Need to make a molecule? Ask this AI for instructions

Baidu shows off its instant pocket translator

The Chinese Internet giant has made significant strides improving machine language translation since 2015, using an advanced form of artificial intelligence known as deep learning, said Hua Wu, the company’s chief scientist focused on natural-language processing. On stage, the Internet-connected device was able to almost instantly translate a short conversation between Wu and senior editor Will Knight. It easily rendered Knight’s questions including ’Where can I buy this device?’

Expressive Speech Synthesis with Tacotron

At Google, we’re excited about the recent rapid progress of neural network-based text-to-speech (TTS) research. In particular, end-to-end architectures, such as the Tacotron systems we announced last year, can both simplify voice building pipelines and produce natural-sounding speech. This will help us build better human-computer interfaces, like conversational assistants, audiobook narration, news readers, or voice design software.

China will publicly shame jaywalkers using facial-recognition technology

The AI company behind the billboards, Intellifusion, is in talks with mobile phone networks and local social media platforms to enforce the new system.

Does my algorithm have a mental-health problem?

Is my car hallucinating? Is the algorithm that runs the police surveillance system in my city paranoid? Marvin the android in Douglas Adams’s Hitchhikers Guide to the Galaxy had a pain in all the diodes down his left-hand side.

AI Cardiologist Aces Its First Medical Exam

When both the AI and expert cardiologists were asked to classify the images, the AI achieved an accuracy of 92 percent. The humans got only 79 percent correct.

Kaggle Public Datasets: 13,000+ Community-Created Datasets

Discover and seamlessly analyze open data

AWS Public Datasets

AWS hosts a variety of public datasets that anyone can access for free. Previously, large datasets such as satellite imagery or genomic data have required hours or days to locate, download, customize, and analyze. When data is made publicly available on AWS, anyone can analyze any volume of data without needing to download or store it themselves.

ProteinNet: A standardized data set for machine learning of protein structure

Guide to Speech Recognition with Python

Far from a being a fad, the overwhelming success of speech-enabled products like Amazon Alexa has proven that some degree of speech support will be an essential aspect of household tech for the foreseeable future. If you think about it, the reasons why are pretty obvious. Incorporating speech recognition into your Python application offers a level of interactivity and accessibility that few technologies can match.

Simplify Adding AI to Your Apps — Core ML Say Hello to Watson

Using Machine Learning to Improve Streaming Quality at Netflix

Network quality is difficult to characterize and predict. While the average bandwidth and round trip time supported by a network are well-known indicators of network quality, other characteristics such as stability and predictability make a big difference when it comes to video streaming. A richer characterization of network quality would prove useful for analyzing networks (for targeting/analyzing product improvements), determining initial video quality and/or adapting video quality throughout playback (more on that below).

North Indian Food or South Indian Food ,Deep Learning Knows it all!

Deploy TensorFlow models

Don’t follow the TensorFlow docs since they explain how to setup a docker image and compile TF serving that takes forever. We can do much better. Some guy made a docker image with everything already compile on it, so we are going to use that one.

Bring Deep Learning Algorithms To Your Security Cameras

AI is quickly revolutionizing the security camera industry. Several manufacturers sell cameras which use deep learning to detect cars, people, and other events. These smart cameras are generally expensive though, compared to their “dumb” counterparts.

The 5 Deep Learning Frameworks Every Serious Machine Learner Should Be Familiar With

Automated front-end development using deep learning

Automated front-end development using deeplearningSketchCode: Go from idea to HTML in 5secondsAshwin Kumar was previously the co-founder of Sway Finance, a Y Combinator-backed startup that used machine learning to automate accounting. At Insight, he developed a model that allows users to create working HTML websites from hand-drawn wireframes, significantly accelerating the design process. He is now a Deep Learning Scientist at Mythic.

Machine Learning Top 10 Open Source Projects

We compared projects with new or major release during this period. Mybridge AI ranks projects based on a variety of factors to measure its quality for professionals.

How AI can learn to generate pictures of cats

The generator is a counterfeiter trying to produce seemingly real data. It has no idea of what the real data is, but it learns to adjust from the feedback of the other model. The discriminator is a inspector trying to determine what the fake counterfeit data is (by comparing it with real data), while trying to not raise false positives on the real data.

GPU-accelerated Neural Networks in JavaScript

While its feature set is geared towards neural networks, deeplearn.js can be described as a general-purpose machine learning framework. Propel is a library for scientific computing that offers automatic differentiation. Gpu.js provides a convenient way to run JavaScript functions on the GPU.

AI Psychosis

Unknown is if they will remain uniquely human. Research in artificial intelligence is making spectacular progress; for manyresearchers, this progress is along the path to developing human-like general AI. This leads to a troubling thought: will a human-like AI inherit human-like disorders of thought?

Robots Are Trying To Pick Strawberries. So Far, They’re Not Very Good At It

Soft robotic fish swims alongside real ones in coral reefs

China tests driverless tanks that could be powered by AI

Understanding Deep Learning through Neuron Deletion

Deep neural networks are composed of many individual neurons, which combine in complex and counterintuitive ways to solve challenging tasks, ranging from machine translation to Go. This complexity grants neural networks their power but also earns them their reputation as confusing and opaque black boxes. Understanding how deep neural networks function is critical for explaining their decisions and enabling us to build more powerful systems.

The next step in Facebook’s AI hardware infrastructure

Machine learning drives many aspects of people’s experience on Facebook. We use automatic language translation systems to remove communication barriers and enable people to interact with each other even if they speak different languages. Our image classification systems not only allow people to search for photos of their favorite moments, but also provide an immersive experience for the visually impaired with “talking images” that can be read with your fingertips.

NGraph: A New Open Source Compiler for Deep Learning Systems

Artificial intelligence can diagnose prostate cancer as well as a pathologist

Splash of Color: Instance Segmentation with Mask R-CNN and TensorFlow

Machine Learning Spots Treasure Trove of Elusive Viruses

Researchers have used artificial intelligence (AI) to discover nearly 6,000 previously unknown species of virus. The work, presented on 15 March at a meeting organized by the US Department of Energy (DOE), illustrates an emerging tool for exploring the enormous, largely unknown diversity of viruses on Earth.

Apps like Waze, Google Maps, and Apple Maps may make traffic conditions worse in some areas, new research suggests

This Hat Can Fool a Face Recognition System Into Thinking You’re Moby

Baidu Apollo Releases Massive Self-driving Dataset; Teams Up With Berkeley DeepDrive

Deciphering China’s AI Dream

This report examines the intersection of two subjects, China and artificial intelligence, both of which are already difficult enough to comprehend on their own. It provides context for China’s AI strategy with respect to past science and technology plans, and it also connects the consistent and new features of China’s AI approach to the drivers of AI development (e.g. hardware, data, and talented scientists). In addition, it benchmarks China’s current AI capabilities by developing a novel index to measure any country’s AI potential and highlights the potential implications of China’s AI dream for issues of AI safety, national security, economic development, and social governance.

VINE: An Open Source Interactive Data Visualization Tool for Neuroevolution

Using Evolutionary AutoML to Discover Neural Network Architectures

The brain has evolved over a long time, from very simple worm brains 500 million years ago to a diversity of modern structures today. The human brain, for example, can accomplish a wide variety of activities, many of them effortlessly — telling whether a visual scene contains animals or buildings feels trivial to us, for example. To perform activities like these, artificial neural networks require careful design by experts over years of difficult research, and typically address one specific task, such as to find what’s in a photograph, to call a genetic variant, or to help diagnose a disease.

AI Detects Papaya Ripeness

Last year, the United States alone imported more than US $107 million worth of fresh papayas as the world’s largest papaya import market. The computer vision software could enable papaya growers to maximize the value of their fruit by sending the ripestpapayas to local markets and saving less ripe papayas for export, says Douglas Fernandes Barbin, a researcher in the department of food engineering at the University of Campinas in São Paulo, Brazil. But he and his colleagues also want to help individual shoppers get their money’s worth in grocery aisles.

Train Your Machine Learning Models on Google’s GPUs for Free

Making music using new sounds generated with machine learning

Technology has always played a role in inspiring musicians in new and creative ways. The guitar amp gave rock musicians a new palette of sounds to play with in the form of feedback and distortion. And the sounds generated by synths helped shape the sound of electronic music.

Kaggle Tensorflow Speech Recognition Challenge

From November 2017 to January 2018 the Google Brain team hosted a speech recognition challenge on Kaggle. The goal of this challenge was to write a program that can correctly identify one of 10 words being spoken in a one-second long audio file. Having just made up my mind to start seriously studying data science with the goal of turning a new corner in my career, I decided to tackle this as my first serious kaggle challenge.

Ten Machine Learning Algorithms You Should Know to Become a Data Scientist

Using AI to match human performance in translating news from Chinese to English

Deep Neural Network implemented in pure SQL over BigQuery

In this post, we’ll implement a deep neural network with one hidden layer (and ReLu and softmax activation functions) purely in SQL. The end-to-end steps for neural network training including the forward pass and back-propagation will be implemented as a single SQL query on BigQuery. As it runs on Bigquery, in effect we are performing distributed neural network training on 100s to 1000s of servers.

Using Google Cloud AutoML to Classify Poisonous Australian Spiders

Google’s new Cloud AutoML Vision is a new machine learning service from Google Cloud that aims to make state of the art machine learning techniques accessible to non-machine learning experts. In this post I will show you how I was able, in just a few hours, to create a custom image classifier that is able to distinguish between different types of poisonous Australian spiders. I didn’t have any data when I started and it only required a very basic understanding of machine learning related concepts.

The Nexus Linking IBM, California Wine, and Climate Modeling

By 2015, Hamann says, the technology—which uses machine learning to extract insights from multiple layers of information—proved itself. Gallo improved yields on the test site while reducing water use. The partnership quickly found another use for IBM’s AI: analyzing a number of variablessuch as proximity to the winery, weather patterns, elevation, days of sunshine, and other factors toidentify suitable locations for new vineyards.

Cracking Open the Black Box of AI with Cell Biology

Deploy Gluon models to AWS DeepLens using a simple Python API

Behind the Motion Photos Technology in Pixel 2

Background motion estimation in motion photos. By using the motion metadata from Gyro and OIS we are able to accurately classify features from the visual analysis into foreground and background.

Open Source Neural Network Synthesizer

NSynth Super is part of an ongoing experiment by Magenta: a research project within Google that explores how machine learning tools can help artists create art and music in new ways. Technology has always played a role in creating new types of sounds that inspire musicians—from the sounds of distortion to the electronic sounds of synths. Today, advances in machine learning and neural networks have opened up new possibilities for sound generation.

How I implemented iPhone X’s FaceID using Deep Learning in Python

A history of machine translation from the Cold War to deep learning

Understanding 2D Dilated Convolution Operation with Examples in Numpy and Tensorflow with…

Paper repro: “Self-Normalizing Neural Networks”

SNNs are really cool: their goal is to create neural networks in which, if the input of any layer is normally distributed, the output will automatically also be normally distributed. This is amazing because normalizing the output of layers is known to be a very efficient way to improve the performance of neural networks, but the current ways to do it (eg BatchNorm) basically involve weird hacks, while in SNN the normalization is an intrinsic part of the mathematics of the neural net.

3 reasons to question the use of emotion-tracking AI in recruiting

PyTorch – Internal Architecture Tour

This post is a tour around the PyTorch codebase, it is meant to be a guide for the architectural design of PyTorch and its internals. My main goal is to provide something useful for those who are interested in understanding what happens beyond the user-facing API and show something new beyond what was already covered in other tutorials.

Semantic Image Segmentation with DeepLab in Tensorflow

Today, we are excited to announce the open source release of our latest and best performing semantic image segmentation model, DeepLab-v3+ [1], implemented in Tensorflow. This release includes DeepLab-v3+ models built on top of a powerful convolutional neural network (CNN) backbone architecture [2, 3] for the most accurate results, intended for server-side deployment. As part of this release, we are additionally sharing our Tensorflow model training and evaluation code, as well as models already pre-trained on the Pascal VOC 2012 and Cityscapes benchmark semantic segmentation tasks.

The Building Blocks of Interpretability

In 2015, our early attempts to visualize how neural networks understand images led to psychedelic images. Soon after, we open sourced our code as DeepDream and it grew into a small art movement producing all sorts of amazing things. But we also continued the original line of research behind DeepDream, trying to address one of the most exciting questions in Deep Learning: how do neural networks do what they do?

The Building Blocks of Interpretability

Interpretability techniques are normally studied in isolation. We explore the powerful interfaces that arise when you combine them — and the rich structure of this combinatorial space.

Google helps Pentagon analyze military drone footage—employees “outraged”

A report from Gizmodo says that Google is partnering with the United States Department of Defense and building drone software. The project will reportedly apply Google’s usual machine learning prowess to identify objects in drone footage. Google’s involvement in the project wasn’t public, but it was apparently discussed internally at Google last week and leaked.

A Virtual Slug Simulator Could Help Researchers Develop Better AI

Ubisoft is using AI to catch bugs in games before devs make them

AI Researchers Aim to Crack Code on ‘Sun Energy’

Since the ’50s, scientists have chased the promise of clean energy from sun-like reactions between deuterium and tritium, the plentiful isotopes of hydrogen. This carbon-free energy, achieved at temperatures of 360 million degrees Fahrenheit, would offer a great way to heat water and, in turn, spin turbines to create countless kilowatts of electricity.

A Framework for Building Artificial Intelligence Capabilities

AI will likely become the most important technology of our era as it’s improved upon over time, but we’re still in the early stages of deployment, CIO Journal Columnist Irving Wladawsky-Berger writes. History shows that even after technologies start crossing over into mainstream markets, it takes considerable time for the new tech to be widely embraced across the economy.

Bonsai AI: Using Simulink for Deep Reinforcement Learning

Simulink provides a great training environment for DRL as it allows 3rd parties like Bonsai to integrate and control simulation models from the outside. This ability is one of the basic requirements for simulation platforms to be feasible for Deep Reinforcement Learning using Bonsai AI. More requirements can be found here.

The Birth of A.I.

Waze famously disrupted GPS navigation by crowdsourcing user data from mobile phones, instead of purchasing costly sensors tied to city infrastructure, as Nokia had done before them. Waze then scaled with low overhead costs by using machine learning algorithms to find precise traffic patterns that optimized each user’s route. The end result of this dynamic was massive layoffs at Nokia, and Google’s acquisition of / integration with Waze in 2013.

How to play Quidditch using the TensorFlow Object Detection API

So what’s new in AI?

I graduated with a degree in AI when the cost of the equivalent computational power to an iPhone was $50 million. A lot has changed but surprisingly much is still the same.

Hacking the Brain with Adversarial Images

This is an example of what’s called an adversarial image: an image specifically designed to fool neural networks into making an incorrect determination about what they’re looking at. Researchers at Google Brain decided to try and figure out whether the same techniques that fool artificial neural networks can also fool the biological neural networks inside of our heads, by developing adversarial images capable of making both computers and humans think that they’re looking at something they aren’t.

Paul Allen Wants to Teach Machines Common Sense

The Microsoft co-founder will give $125 million to his nonprofit research lab to help develop technology that adds common sense to artificial intelligence.

Machines Stomp Lawyers in Legalese Smackdown

Cue the sad tuba and attorney jokes: Machines just landed the hurt on lawyers. LawGeex, an Israel-based startup focused on automating contract reviews, released a study showing its AI software pummels lawyers in document review accuracy. The AI service outperformed 20 corporate lawyers at identifying legal risks in nondisclosure agreement contracts.

Humanizing Medicine with AI – Enabling Docs to Speak with Their Patients

Understanding word vectors in NLP

In this tutorial, I’m going to show you how word vectors work. This tutorial assumes a good amount of Python knowledge, but even if you’re not a Python expert, you should be able to follow along and make small changes to the examples without too much trouble.

AI Cheats at Old Atari Games by Finding Unknown Bugs in the Code

Mobile Real-time Video Segmentation

We use big convolution kernels with large strides of four and above to detect object features on the high-resolution RGB input frame. Convolutions for layers with a small number of channels (as it is the case for the RGB input) are comparably cheap, so using big kernels here has almost no effect on the computational costs.

Can AI Ever Learn To Follow Its Gut?

Academics, economists, and AI researchers often undervalue the role of intuition in science. Here’s why they’re wrong.