Recent developments in Artificial Intelligence (AI) are showing just how far we have come with the technology. Take for example the recent showcase from Google, where a telephone call is made by Google Assistant to book a haircut. The technology is now able to understand the nuances of conversation and is quite astonishing at first watch. With this ability to save time and cut admin, it is no wonder that adoption rates of AI in businesses have grown 60 percent in the past year alone. However, AI needs to be programmed by a human and in today's working environment, this can be problematic.

The development of Artificial General Intelligence (AGI) promises to be a major event. Along with its many potential benefits, it also raises serious safety concerns (Bostrom, 2014). The intention of this paper is to provide an easily accessible and up-to-date collection of references for the emerging field of AGI safety. A significant number of safety problems for AGI have been identified. We list these, and survey recent research on solving them. We also cover works on how best to think of AGI from the limited knowledge we have today, predictions for when AGI will first be created, and what will happen after its creation. Finally, we review the current public policy on AGI.

I give nothing as duties; What others give as duties, I give as living impulses. Abstract Walt Whitman Various moral conundrums plague population ethics: The Non-Identity Problem, The Procreation Asymmetry, The Repugnant Conclusion, and more. I argue that the aforementioned moral conundrums have a structure neatly accounted for, and solved by, some ideas in computability theory. I introduce a mathematical model based on computability theory and show how previous arguments pertaining to these conundrums fit into the model. This paper proceeds as follows. First, I do a very brief survey of the history of computability theory in moral philosophy. Second, I follow various papers, and show how their arguments fit into, or don't fit into, our model. Third, I discuss the implications of our model to the question why the human race should or should not continue to exist. Finally, I show that our model ineluctably leads us to a Confucian moral principle. In 1931, Gödel introduced his Incompleteness Theorem. The results showed that, roughly, consistency and completeness cannot coexist in a formal system. In Gödel, Escher, Bach, the most seminal treatment on the topic thus far, Hofstadter puts it this way: "for any record player, there are records which it cannot play because they will cause its indirect self-destruction" There is an inkling of intuition here that may be grasped, but the precise mathematical idea is difficult to understand, and Hofstadter's explanation cannot be said to be precise or thorough. Rather than explain the theorem in detail, however, I want to explain how the theorem has been applied, or complained about, so that the reader can build a better intuition about it. The theorem has had a sizable impact in academic philosophy, most notably in the form of arguments against determinism, and as arguments against strong AI. However, as far as the author's knowledge goes, its implications have not been milked to their full potential in academic moral philosophy. The theorem has certainly been discussed in moral philosophy, however, such as in the following speech by the British philosopher J. R. Lucas [Luc98]: Moral and political philosophy will be different once reason is allowed to regain its ancient sway.... Although the way Gödel's theorem was proved follows a somewhat standard route, the upshot of the proof is that reasoning, even mathematical reasoning, By doing that it becomes free.

Just over fifty years ago, a French journalist, Jean-Jacques Servan-Schreiber, published his book, Le Défi Américain (aka The American Challenge, 1967). It presented the United States and Europe as engaged in a silent economic war. In that war, he wrote, Europe was being completely outclassed on all fronts in dealing with the Third Industrial Revolution (electronics, information technology, and automation). The invading industrial armies of the day--1960s giants such as General Motors and IBM--were becoming dominant in Europe because of stronger and more flexible management techniques, technological tools, and research capacity. The book became an international hit, selling an unprecedented 600,000 copies in France alone.

If you're grasping for the deeper meaning of an essay or article, consider the possibility that it may not be in the words themselves, but hidden in the shape of the letters. It really could be the case, now that researchers from Columbia University have developed a method called FontCode, which plants data in text through tiny changes in how the letters are shaped. The method is a steganographic technique, meaning it hides secret information in plain sight such that only its intended recipient knows where to look for it and how to extract it. FontCode can be applied to hundreds of common fonts, like Helvetica or Times New Roman, and works in word processors like Microsoft Word. Data encoded with FontCode can also endure across any image-preserving digital format, like PDF or PNG. The secret data won't persist after, say, copy and pasting FontCode text between text editors.

It's hard to predict how Artificial Intelligence (AI) will behave. In 2016, Microsoft released Tay, an AI Twitter bot designed to interact with people through conversational language and get smarter along the way by observing, learning and mimicking. The designers perhaps underestimated the Twitterati's penchant for mischief. Soon after Tay was launched, people started sending it racist, misogynistic and hateful tweets. And Tay, true to design, absorbed and internalised this behaviour, and began tweeting similar nasty sentiments back at them. Microsoft had to scramble and work overtime to delete the offensive tweets.

Machine learning systems trained to minimize prediction error may often exhibit discriminatory behavior based on sensitive characteristics such as race and gender. One reason could be due to historical bias in the data. In various application domains including lending, hiring, criminal justice, and advertising, machine learning has been criticized for its potential to harm historically underrepresented or disadvantaged groups. In this post, we talk about our recent work on aligning decisions made by machine learning with long term social welfare goals. Commonly, machine learning models produce a score that summarizes information about an individual in order to make decisions about them.

In Pyrrho Investments v MWB Business Exchange, Master Paul Matthews of the Chancery division supported the use of software in scoring documents for relevance. He found there was no evidence that software would be less accurate than manual review and keyword searches. He added that software could provide greater consistency in searching more than 3 million documents that could be involved in the disclosure. A final reason was that both sides had agreed to the use of the software, which would be much cheaper than a manual search – they just wanted the court's approval. However, in May, the High Court went further when two undisclosed parties disagreed on whether predictive coding software should be used.

In Pyrrho Investments v MWB Business Exchange, Master Paul Matthews of the Chancery division supported the use of software in scoring documents for relevance. He found there was no evidence that software would be less accurate than manual review and keyword searches. He added that software could provide greater consistency in searching more than 3 million documents that could be involved in the disclosure. A final reason was that both sides had agreed to the use of the software, which would be much cheaper than a manual search – they just wanted the court's approval. However, in May, the High Court went further when two undisclosed parties disagreed on whether predictive coding software should be used.

Much of the attention has focused on Theranos founder Elizabeth Holmes. But another character played a central role behind the scenes in the alleged fraud: Ms. Holmes's boyfriend, Ramesh "Sunny" Balwani, according to more than three dozen former Theranos employees who interacted with Mr. Balwani extensively over a number of years. Mr. Balwani, who met Ms. Holmes when she was a teenager, jointly ran the company with her for seven years as president and chief operating officer and enforced a corporate culture of secrecy and fear until his departure in the spring of 2016, the former employees say. Unlike Ms. Holmes and Theranos, who reached a settlement with the SEC to resolve the agency's civil charges in March without admitting or denying wrongdoing, Mr. Balwani has denied separate charges the SEC filed against him in a parallel action and is fighting them in a California federal court. A spokeswoman for Mr. Balwani provided a statement from his lawyer, Jeffrey B. Coopersmith, saying Mr. Balwani accurately represented Theranos to investors to the best of his ability, worked hard to maximize shareholder value and took on significant risk investing in the company while never benefiting financially from his work.