Will progress in Artificial Intelligence provide humanity with a boost of unprecedented strength to realize a better future, or could it present a threat to the very basis of human civilization? The future of artificial intelligence is up for debate, and the Origins Project is bringing together a distinguished panel of experts, intellectuals and public figures to discuss who's in control. Eric Horvitz, Jaan Tallinn, Kathleen Fisher and Subbarao Kambhampati join Origins Project director Lawrence Krauss. Recorded Saturday, February 25th, 2017 Eric Horvitz is managing director of Microsoft Research's main Redmond Lab, an American computer scientist, and technical fellow at Microsoft. Horvitz received his PhD and MD degrees at Stanford University, and has continued his research and work in areas that span theoretical and practical challenges of machine learning and inference, human-computer interaction, artificial intelligence, and more.

The workplace is going to look drastically different ten years from now. The coming of the Second Machine Age is quickly bringing massive changes along with it. Manual jobs, such as lorry driving or house building are being replaced by robotic automation, and accountants, lawyers, doctors and financial advisers are being supplemented and replaced by high level artificial intelligence (AI) systems. So what do we need to learn today about the jobs of tomorrow? The robots and computers of the future will be based on a degree of complexity that will be impossible to teach to the general population in a few short years of compulsory education.

The probability that a user will click a search result depends both on its relevance and its position on the results page. The position based model explains this behavior by ascribing to every item an attraction probability, and to every position an examination probability. To be clicked, a result must be both attractive and examined. The probabilities of an item-position pair being clicked thus form the entries of a rank-$1$ matrix. We propose the learning problem of a Bernoulli rank-$1$ bandit where at each step, the learning agent chooses a pair of row and column arms, and receives the product of their Bernoulli-distributed values as a reward. This is a special case of the stochastic rank-$1$ bandit problem considered in recent work that proposed an elimination based algorithm Rank1Elim, and showed that Rank1Elim's regret scales linearly with the number of rows and columns on "benign" instances. These are the instances where the minimum of the average row and column rewards $\mu$ is bounded away from zero. The issue with Rank1Elim is that it fails to be competitive with straightforward bandit strategies as $\mu \rightarrow 0$. In this paper we propose Rank1ElimKL which simply replaces the (crude) confidence intervals of Rank1Elim with confidence intervals based on Kullback-Leibler (KL) divergences, and with the help of a novel result concerning the scaling of KL divergences we prove that with this change, our algorithm will be competitive no matter the value of $\mu$. Experiments with synthetic data confirm that on benign instances the performance of Rank1ElimKL is significantly better than that of even Rank1Elim, while experiments with models derived from real data confirm that the improvements are significant across the board, regardless of whether the data is benign or not.

We met Tanmay while he was giving a keynote speech to our tech associates (yes, he gives keynote speeches around the world). Please enjoy Tanmay's article on 5 machine learning services powered by Watson. Computers have never been good at understanding non-mathematical language, but a branch of artificial intelligence aims to change that. Machine learning (ML), a subset of artificial intelligence, aims to enable computers to understand natural language and allow them to reason for themselves. Leading this initiative are services like IBM's Watson, an open platform accessible to developers around the world. In today's world, everyone is using machine learning technologies on a daily basis, whether they realize it or not.

OpenAI research has a short introduction on their newest research "Learning to Communicate". There are many trends that I watch for in the field of Deep Learning. Two trends that are related and I believe going to be very promising areas are language learning and multi-agent communication. If you have not been watching, this week has had a tremendous release of papers involving the former and culminating with OpenAI's post, stitching it all together! Let me explain though what transpired in this amazing week.

Forty-three high school seniors who took part in an exam-taking competition in Chengdu, China, officially outperformed the math robot they were up against. The robot, developed by China's Chengdu Zhun Xing Yun Xue Technology Co. received an average score of 93, while the class received an average score of 106, the Huffington Post reports. The AI score is on par with the company's estimate of between 90 and 110. According to the Chengdu Economic Daily, the AI and the students, a liberal arts class from Shishi Tianfu High School, were all required to complete the test in a two-hour timeframe. "I think although it's a math test, we won it because of our literary skills," She Yujia, a student participant who scored 135 on the exam, told HuffPo.

Before Kanye West gets to the White House, first, we'll have to survive the robot apocalypse brought about by his A.I.-powered doppelgänger. It's a very real piece of software created by a high school student from West Virginia. You can use Alexa in Amazon's app now, and it's really smart Robbie Barrat, a 17-year-old hip-hop fan and coding whiz, taught himself to code using open source software, according to a report from Quartz. Initially, the software simply rearranged 6,000 Kanye rap phrases to create new songs, but now the software has been modified to create original rap lines using the Kanye word bank. On the YouTube Page demonstrating the software's ability, Barrat says, "Excluding the beat; this song was written 100 percent by a deep neural network."

With Safari, you learn the way you learn best. Get unlimited access to videos, live online training, learning paths, books, tutorials, and more.

As Quartz reports, a West Virginia teenager has created an artificial intelligence emcee whose style is based entirely on Kanye West's body of work. "All of the sudden I had a week to make a neural network that could rap," Barrat explained to Quartz. A self-taught programmer, Barrat cooked up most of the code for his digital lyricist in the course of one afternoon; it only took him a few extra days to complete the effort. The West Virginia teen's AI rapper takes its cues from 6,000 different Kanye West lines and is able to produce its own bars and flows like a traditional artist would. "Originally it just rearranged existing rap lyrics, but now it can actually write word-by-word," Barrat says.