Elon Musk, recently busying himself with calling people "pedo" on Twitter and potentially violating US securities law with what was perhaps just a joke about weed – both perfectly normal activities – is now involved in a move to terrify us all. The non-profit he backs, OpenAI, has developed an AI system so good it had me quaking in my trainers when it was fed an article of mine and wrote an extension of it that was a perfect act of journalistic ventriloquism. As my colleague Alex Hern wrote yesterday: "The system [GPT2] is pushing the boundaries of what was thought possible, both in terms of the quality of the output, and the wide variety of potential uses." GPT2 is so efficient that the full research is not being released publicly yet because of the risk of misuse. And that's the thing – this AI has the potential to absolutely devastate.

The great potential of artificial intelligence (AI) comes with some dire warnings, both real and imagined. With the goal of applying AI technologies to achieve extraordinary good, Noodle.ai is presenting the good and bad side in a new infographic. With such an ominous title, "AI The Dark Side versus The Force for Good" you would think the bad news is worse than it is, but the infographic is cautiously optimistic about the technology. This bodes well for small businesses who are looking to implement AI as part of their daily operations. And as more small businesses increase their digital presence, the benefits of AI will be more impactful for the organization as whole.

AI or ML is thus inserted in a complete array of cyber measures called Intelligent Security Solutions. These are the software, protocols or raw code, which is added to the IT system of a company or an organization. Thereafter, another layer of security is added to the traditional safety protocols, which has an ability to learn from threats, security breaches and other data collected through their mechanisms. Thus, AI works through thousands of data and learns from the same. However, whatever works for cybersecurity experts also works for the hackers. All the data spread out there on the internet, can be used from the other side of the cybersecurity spectrum. This is how, security breaches could be prevented by AI-powered systems through the data. Hackers have always tried to overcome anti-malware software, antivirus, firewalls and other security barriers, and for doing so, they have imagined all ways to blend and escape themselves within the system without being recognized. But, AI has enormously multiplied the possibilities.

Welcome to California Inc., the weekly newsletter of the L.A. Times Business Section. He's the former Enron chief executive whose scheming helped bankrupt PG&E during the California energy crisis. On Friday, Skilling was released from federal custody after serving 12 years in prison for multiple counts of securities fraud, conspiracy and other crimes. Home prices: The latest Case-Shiller home price index comes out Tuesday. In November, seasonally adjusted home prices for the benchmark 20-city index were up 0.3% month over month.

Artificial Intelligence is making its way into various organizations and corporations. But still, the discussion that whether AI is a boon or a ban is a never-ending one. Despite the fact that the AI program was written in 1958, it has seen tremendous growth in recent years. We have seen Sophia, Alexa voice bot and many other similar examples that are making people wonder how is this field can turn against humans. Experts are researching ways to add AI into different domains, and cybersecurity is one of them.

A storm is brewing over a new language model, built by non-profit artificial intelligence research company OpenAI, which it says is so good at generating convincing, well-written text that it's worried about potential abuse. That's angered some in the community, who have accused the company of reneging on a promise not to close off its research. OpenAI said its new natural language model, GPT-2, was trained to predict the next word in a sample of 40 gigabytes of internet text. The end result was the system generating text that "adapts to the style and content of the conditioning text," allowing the user to "generate realistic and coherent continuations about a topic of their choosing." The model is a vast improvement on the first version by producing longer text with greater coherence.

Speaking on the record to an invited audience at RUSI on 21 January 2019, GCHQ Deputy Director for Strategic Policy Paul Killworth described how Artificial Intelligence (AI) and Machine Learning (ML) have the potential to improve the effectiveness and efficiency of various intelligence functions. However, these capabilities bring with them complex legal and ethical considerations, and there is a strong public expectation that the UK's intelligence agencies will act in a way that protects citizens' rights and freedoms. The national security community has expressed a desire to engage in a more open dialogue on these issues, with Killworth stressing that'it is absolutely essential that we have the debates around AI and machine learning in the national security space that will deliver the answers and approaches that will give us public consent'. However, it may prove difficult to provide sufficient reassurances to the public concerning national security uses of AI, due to understandably high levels of sensitivity. Public acceptance of intelligence agencies' use of technology is driven by two conflicting sentiments. On the one hand, there is a high expectation that the national security community will protect citizens from threats to their safety, and adopt new methods that may allow them to do this more effectively.

Binscatter is very popular in applied microeconomics. It provides a flexible, yet parsimonious way of visualizing and summarizing large data sets in regression settings, and it is often used for informal evaluation of substantive hypotheses such as linearity or monotonicity of the regression function. This paper presents a foundational, thorough analysis of binscatter: we give an array of theoretical and practical results that aid both in understanding current practices (i.e., their validity or lack thereof) and in offering theory-based guidance for future applications. Our main results include principled number of bins selection, confidence intervals and bands, hypothesis tests for parametric and shape restrictions of the regression function, and several other new methods, applicable to canonical binscatter as well as higher-order polynomial, covariate-adjusted and smoothness-restricted extensions thereof. In particular, we highlight important methodological problems related to covariate adjustment methods used in current practice. We also discuss extensions to clustered data. Our results are illustrated with simulated and real data throughout. Companion general-purpose software packages for \texttt{Stata} and \texttt{R} are provided. Finally, from a technical perspective, new theoretical results for partitioning-based series estimation are obtained that may be of independent interest.

The use of convolutional neural networks has led to tremendous achievements since Krizhevsky et al. [1] presented AlexNet in 2012. Despite efforts to understand the inner workings of such neural networks, they mostly remain black boxes that are hard to interpret or explain. The issue was exaggerated in 2013 when Szegedy et al. [2] showed that "adversarial examples" - images perturbed in such a way that they fool a neural network - prove that neural networks do not simply work correctly the way one might naïvely expect. Typically,such adversarial attacks change an input only slightly, but in an adversarial manner, such that humans would not regard the difference of the inputs relevant, but machines do. There are various types of attacks, such as one pixel attacks, attacks that work in the physical world, and attacks that produce inputs fooling several different neural networks without explicit knowledge of those networks [3, 4, 5].