To describe Max Tegmark's career as "storied" is to do the Swedish-American physicist a disservice. He's published more than 200 publications and developed data analysis tools for microwave background experiments. And he's been elected as a Fellow of the American Physical Society for his contributions to cosmology. In 2015, Elon Musk donated $10 million to FLI to advance research into the ethical, legal, and economic effects of AI systems. Tegmark's latest book, Life 3.0: Being Human in the Age of Artificial Intelligence, postulates that neural networks of the future may be able to redesign their own hardware and internal structure.

Voices for the need of an ethical Artificial Intelligence are getting louder. AI has a great many applications, and has been solving various problems for humans. But such a powerful technology raises equal concerns about its possible misuse. And there have already been multiple cases of AI being used for malicious purposes. It is an advanced AI software that creates an artificial version of a person's face accurately matching that of another.

Google CEO Sundar Pichai yesterday published his company's new rules governing the development of AI. Over the course of seven principles he lays out a broad (and useless) policy leaving more wiggle room than a pair of clown pants. If you tell the story of Google's involvement in building AI for the US military backwards it makes perfect sense. In such a case, the tale would begin with the Mountain View company creating a policy for developing AI, and then it would use those principles to guide its actions. Unfortunately the reality is the company has been developing AI for as long as it's been around.

Attacks are getting more complex. This is especially true when it comes to cyberwar, so much so that government sponsored attacks have been bolstered by research investments that approach military proportions. Just look at the recent report published by the US State Department, which said that strategies for stopping cyber attacks need to be fundamentally reconsidered in light of complex cyber threats posed by rival states. In order to detect and stop these attacks, innovation is required. I say that because anomaly detection based on traditional correlation rules often results in too many false positives and events that can reasonably be manually reviewed.

Donald Trump's chief technology advisor, Michael Kratsios, said that the US government would release any data that might help fuel AI research in the United States, although he didn't specify immediately what kind of data would be released, nor who would be eligible to receive the information. Speaking at EmTech Next, a conference held at MIT this week, Kratsios, deputy assistant to the President and deputy US chief technology officer, said that the government is looking for ways to open up federal data to AI researchers. "Anything that we can do to unlock government data we're committed to," Kratsios told MIT Technology Review. "We'd love to hear from any academic that has any insights." Data has been a key factor behind recent advances in artificial intelligence.

Getty Close-up adult hand typing on laptop A centre for artificial intelligence and public policy is looking to address the unique issues New Zealand will face, and is currently facing, in the era of AI. The centre has been launched in Otago and will explore policy options for managing the introduction of technologies, to maximise their benefits and minimise potential harms. Co-director of the centre, Professor James Maclaurin, said New Zealand's size sets it apart from other countries and it is important to have people acting in an advisory role. "Europe has just passed its general data protection regulations but it is a very big player so if they pass laws Facebook and Google really have to listen to them. "New Zealand is a different environment."

Have you noticed that AI has finally passed into everyday reality? No one needs to reference science fiction apologetically to talk about it. And now that it's real, our questions about it need to get real in response. No, I don't mean, "Will the robots rise up and take over?" Rather, now that AI is growing up and gaining power, our questions have to grow up with it.

The statement comes amid employee discontent over Google's involvement in Project Maven, a controversial Pentagon AI program that seeks, among other things, to use machine learning to quickly detect and categorize objects in images captured by drones. After several employees quit and thousands of others signed a petition against the program, Google has said it will end its participation when its current contract expires next year. Other tech companies have been closemouthed about their role in the program. In a blog post, Google CEO Sundar Pichai said the company won't stop working with the military entirely: It will still potentially work with the armed forces on areas including cybersecurity, recruitment and training, veterans' healthcare and search and rescue. Google is widely seen as a potential contender for a massive contract to move Defense Department systems to cloud servers.

LexNLP is an open source Python package focused on natural language processing and machine learning for legal and regulatory text. The package includes functionality to (i) segment documents, (ii) identify key text such as titles and section headings, (iii) extract over eighteen types of structured information like distances and dates, (iv) extract named entities such as companies and geopolitical entities, (v) transform text into features for model training, and (vi) build unsupervised and supervised models such as word embedding or tagging models. LexNLP includes pre-trained models based on thousands of unit tests drawn from real documents available from the SEC EDGAR database as well as various judicial and regulatory proceedings. Keywords: natural language processing, legal, regulatory, machine learning, segmentation, extraction, open source, Python 1. Introduction Over the last two decades, many high-quality, open source packages for natural language processing and machine learning have been released. Researchers and developers can quickly write applications in languages such as Java, Python, and R that stand on the shoulders of comprehensive, well-tested libraries like Stanford NLP ([1]), OpenNLP ([2]), NLTK ([3]), spaCy ([4]), scikit-learn ([5], [6]), Weka ([7]), and gensim ([8]).

Diffusions and related random walk procedures are of central importance in many areas of machine learning, data analysis, and applied mathematics. Because they spread mass agnostically at each step in an iterative manner, they can sometimes spread mass "too aggressively," thereby failing to find the "right" clusters. We introduce a novel Capacity Releasing Diffusion (CRD) Process, which is both faster and stays more local than the classical spectral diffusion process. As an application, we use our CRD Process to develop an improved local algorithm for graph clustering. Our local graph clustering method can find local clusters in a model of clustering where one begins the CRD Process in a cluster whose vertices are connected better internally than externally by an $O(\log^2 n)$ factor, where $n$ is the number of nodes in the cluster. Thus, our CRD Process is the first local graph clustering algorithm that is not subject to the well-known quadratic Cheeger barrier. Our result requires a certain smoothness condition, which we expect to be an artifact of our analysis. Our empirical evaluation demonstrates improved results, in particular for realistic social graphs where there are moderately good---but not very good---clusters.