Is Artificial Intelligence Finally Coming into Its Own?


In March the company bought a startup cofounded by Geoffrey Hinton, a University of Toronto computer science professor who was part of the team that won the Merck contest. Extending deep learning into applications beyond speech and image recognition will require more conceptual and software breakthroughs, not to mention many more advances in processing power. Programmers would train a neural network to detect an object or phoneme by blitzing the network with digitized versions of images containing those objects or sound waves containing those phonemes. A team led by Stanford computer science professor Andrew Ng and Google Fellow Jeff Dean showed the system images from 10 million randomly selected YouTube videos.

Homeless, assaulted, broke: drivers left behind as Uber promises change at the top

The Guardian

The answer: Travis Kalanick, Uber's 40-year-old co-founder and chief executive, was taking a leave of absence from the taxi-hailing app he has transformed into a global behemoth valued at almost $70bn. A week before the incident, Patterson had signed a lease through Uber's vehicle loan program, known as Xchange. Unable to keep up with her bills, she ended up living out of her leased Uber vehicle for about a month until, running of options, she moved back home to Chicago, where she fell further into debt before her car was repossessed. Though the lawsuit and departure of Levandowski could hamper Uber's development of its own technology, the company has continued testing autonomous cars with pilot projects in Pittsburgh, Arizona and San Francisco, where Uber's self-driving vehicles were caught driving through red lights and initially banned by regulators.

Samsung's Bixby finally gets a voice -- sort of


One of the most anticipated new features of the Samsung Galaxy S8 and S8 prior to the phone's launch in March was an artificial intelligence assistant named Bixby that you were supposed to be able to control by voice. Unfortunately while some of Bixby's capabilities made into onto the phones, the voice-based commands that would make Bixby respond more like the Google Assistant, Apple's Siri, Microsoft's Cortana and Amazon's Alexa was delayed, at least in the U.S. (Bixby is fully operational in South Korea, where Samsung is based). On Friday, Samsung Electronics America announced it will give "select" Galaxy S8 and S8 users early access to Bixby's vocal capabilities as part of what it still considered an early preview test. Indeed, Samsung thinks of Bixby as an intelligent user assistant that relies on other senses.

AI Could Target Autism Before It Even Emerges--But It's No Cure-All


By training a machine learning algorithm on their behavior and earlier MRI data, the scientists built a model that predicted 9 of those 11 autism cases, with no false positives. Right now, researchers tracking autism development focus on infant siblings of people with autism; they have 1 in 5 chance of developing autism, compared to around 1 in 100 for the general population. But current autism therapies for babies and toddlers focus on their specific behavioral deficits--teaching children to communicate needs, to play with toys, and to have positive interactions with caregivers. The UNC group's next goal is to predict specific autism symptoms, correlating brain scans with future language difficulties, sensory sensitivities, social difficulties, or repetitive behaviors.

Can my computer recognise my cat?


In 2012, Google created an'artificial neural network' and fed it millions of pictures from the internet. Could your computer now identify your cat more accurately than you?

An Artificial Intelligence Developed Its Own Non-Human Language


In the report, researchers at the Facebook Artificial Intelligence Research lab describe using machine learning to train their "dialog agents" to negotiate. At one point, the researchers write, they had to tweak one of their models because otherwise the bot-to-bot conversation "led to divergence from human language as the agents developed their own language for negotiating." In other words, the model that allowed two bots to have a conversation--and use machine learning to constantly iterate strategies for that conversation along the way--led to those bots communicating in their own non-human language. Already, there's a good deal of guesswork involved in machine learning research, which often involves feeding a neural net a huge pile of data then examining the output to try to understand how the machine thinks.

AI bots are learning to team up by wrangling digital swine in Minecraft


This much seems clear from a contest organized by Microsoft researchers to test how artificially intelligent agents could cooperate to solve tricky problems. For the Microsoft contest, AI agents worked together inside Project Malmo, a special version of the open-ended computer game Minecraft. The top teams in the Malmo Collaborative AI Challenge used cutting-edge machine-learning approaches such as deep learning to train their agents to work together. Pedro Domingos, a professor at the University of Washington who studies machine learning and data mining, says training AI software inside simulated environments has its drawbacks.

Stanford Robots Load a Hovering Drone, Solve a Marble Maze, and More

IEEE Spectrum Robotics Channel

Robots that sketch, play ping pong, solve mazes, and attempt to juggle strutted their stuff at the annual demo day for Stanford's Experimental Robotics class. Each year, Professor Oussama Khatib's students aim to teach a selection of industrial robots some new skills. The toughest project: teaching a Kuka robot to juggle. The other projects included two robots that had been taught to draw (a Sawyer and a Puma), a Ping Pong–playing robot (the Kuka again) that scored a few points against its human opponents, a cowboy hat–wearing Sawyer robot that shot at a moving target, a Puma 500 robot that manipulated a maze to send a ball along the correct path, and a drone-loading Sawyer robot that tracked a less-than-stable hovering drone.

Man vs. Machine: Robot Calls Police After Being Attacked By Drunk Man

International Business Times

A drunk man reportedly ran into an armless K5 robot in the Knightscope parking lot in Mountain View, California and met his match. The company spokesman Stacy Dean Stephens said that members of the robot company Knightscope -- which developed the robot that appears similar to the iconic Star Wars Droid R2D2 -- came out and detained Sylvain as the police came. Device Warns It's Wearer When Robots Impersonate Humans Knightscope is a company based in Silicon Valley and develops fully autonomous robots, known as the K5 Beta Prototype and has been used for security purposes in schools, businesses, and neighborhoods. The company developed the robot security guards in the wake of the Sandy Hook Elementary School massacre when 20-year-old Adam Lanza fatally shot 20 children between six and seven years old before committing suicide when the police arrived.

DAVID BRIN: How Might Artificial Intelligence Come About?


Those fretfully debating artificial intelligence (AI) might best start by appraising the half dozen general pathways under exploration in laboratories around the world. Some of these endeavors start with mathematical theories that posit Artificial General Intelligence (AGI) on infinitely-powerful machines, then scale down.