Today Silicon Valley chip maker Movidius released the Fathom Neural Compute Stick. It looks like a measly thumb drive, but inside it packs a high-end visual processing unit that can do a bunch of advanced image recognition. That chip, which is called the Myriad 2, is the same one powering the computer vision and autonomous features in DJI's latest drone. The Fathom is basically a plug-and-play version of the Myriad 2, and Movidius hopes engineers will use it to build deep learning features like like pixel-by-pixel imagine labeling and advanced video analytics into their existing products. "It lets you implement machine learning in an ad hoc manner," Cormac Brick, head of machine learning at Movidius, tells The Verge.

High-tech entrepreneur Elon Musk has launched an open-source training "gym" for artificial-intelligence programmers. It's an interesting move for a man who in 2014 said artificial intelligence, or A.I., will pose a threat to the human race. "I think we should be very careful about artificial intelligence," Musk said about a year and a half ago during an MIT symposium. "If I were to guess at what our biggest existential threat is, it's probably that... with artificial intelligence, we are summoning the demon. In all those stories with the guy with the pentagram and the holy water, and he's sure he can control the demon. Today, Musk is moving to help programmers use A.I. and machine learning to build smart robots and smart devices. "We're releasing the public beta of OpenAI Gym, a toolkit for developing and comparing reinforcement learning (RL) algorithms," wrote Greg Brockman, OpenAI's CTO, and John Schulman, a scientist working with OpenAI, in a blog post . "We originally built OpenAI Gym as a tool to accelerate our own RL research.

Piano Tiles 2 may drive most users up the wall as it becomes increasingly difficult to keep up with the speed, but for a robot, smashing human records is no big deal. The popular game works for smartphone or tablet and requires you'don't tap the white,' pressing only the black tiles as they cascade down the screen. In a recent video posted to YouTube, a homemade robot proves is not only capable of playing the game, but hits a record-breaking 21.079 tiles per second before missing a step. In a recent video posted to YouTube, a robot proves is not only capable of playing the game, but hits a record-breaking 21.079 tiles per second before missing a step. According to the achievement screen which pops up after the robot stops playing, it'beat 100% players globally,' and set a new personal best The first game mastered by a computer was noughts and crosses (also known as tic-tac-toe) in 1952.

A new tech startup is merging deep learning with autonomous cars. Drive.ai has become the 13th company to be granted a license to test autonomous cars on public roads in California. If you haven't yet heard of them, you're not alone. Drive.ai has been in stealth mode for the past year. However, the company recently closed 12 million in Series A funding while developing deep learning technologies with their team of experts who specialize in everything from natural language processing, computer vision, and autonomous driving.

Recurrent neural networks (RNN) are simple dynamical systems whose computational power has been attributed to their short-term memory. Short-term memory of RNNs has been previously studied analytically only for the case of orthogonal networks, and only under annealed approximation, and uncorrelated input. Here for the first time, we present an exact solution to the memory capacity and the task-solving performance as a function of the structure of a given network instance, enabling direct determination of the function–structure relation in RNNs. We calculate the memory capacity for arbitrary networks with exponentially correlated input and further related it to the performance of the system on signal processing tasks in a supervised learning setup. We compute the expected error and the worst-case error bound as a function of the spectra of the network and the correlation structure of its inputs and outputs.

Google's DeepMind artificial intelligence (AI) research group today announced that for all future research it will use TensorFlow, a machine learning library that Google open-sourced last year, instead of Torch, an older framework. The move suggests that some of Google's brightest AI minds are convinced of the promise of Google's own open source software; TensorFlow is now good enough for DeepMind. "We believe that TensorFlow will enable us to execute our ambitious research goals at much larger scale and an even faster pace, providing us with a unique opportunity to further accelerate our research programme," Koray Kavukcuoglu, a research scientist at Google DeepMind and one of Torch's core contributors, wrote in a blog post. This is important because of DeepMind's considerable capabilities -- earlier this year its AlphaGo AI player of the ancient Chinese board game Go beat top-ranked Go player Lee Sedol. To be sure, DeepMind is not Google's only AI research unit.

When it comes to the potential impact AI could bring, mass-unemployment is probably a more realistic concern for us than, say, the Skynet (the murderous AI system of the Terminator film franchise), says Martin Ford, a technology entrepreneur and author of two books about how tomorrow's technology might give a fatal blow to the social structure that we thrive on today. If we look far enough into the future, Ford says, few jobs would be safe from being automated, as algorithms with deep learning capabilities would take over not only entry-level jobs, but also those requiring years of training and experience. "In terms of jobs [that may be done by AI]… the important word there is'predictable'," Ford says. "If another smart person could study a record of everything you've done in the past in your job and based on that, learn how to do your job, then someday, maybe a machine might be able to do the same thing." Ford's warning of a jobless future is not entirely new; and as always, the idea is controversial because opponents argue that historically, workers have survived rounds of technological revolution and they always managed to find other jobs in newly emerged industries.

There is now a technique that links deep learning software and a microscope; it is now easier than ever to pinpoint cancer cells. It can be very difficult to identify cancer purely by using blood samples and while there is an age old system of adding chemicals to the blood to make it easier, it then ruins that sample about any other form of tests. The abnormal structure can be used, and while useful, this takes longer, and it is also possible to identify a healthy cell as one that contains cancer. The device that invented by UCLA professor uses deep learning and photonic time stretches to analyze 36 million images per second. The microscope involved is called a photonic time stretch microscope and works by breaking nanosecond long light pulses into lines so that they can be entered into a computer.

It's no secret that Google has broad ambitions in healthcare. But a document obtained by New Scientist reveals that the tech giant's collaboration with the UK's National Health Service goes far beyond what has been publicly announced. The document – a data-sharing agreement between Google-owned artificial intelligence company DeepMind and the Royal Free NHS Trust – gives the clearest picture yet of what the company is doing and what sensitive data it now has access to. The agreement gives DeepMind access to a wide range of healthcare data on the 1.6 million patients who pass through three London hospitals run by the Royal Free NHS Trust – Barnet, Chase Farm and the Royal Free – each year. This will include information about people who are HIV-positive, for instance, as well as details of drug overdoses and abortions. The agreement also includes access to patient data from the last five years.

Artificial intelligence (AI) is the convenient future. It is one of the most promising and transformative opportunities of our time. We are closer to the near future where virtual assistants, bots, and software agents will act more and more like people. Some the biggest advances in AI are being developed inside tech giants such as Google (Deep Mind) and IBM (Watson). But there are still a lot of great opportunities for young startups to explore.