Facebook's tech boss on how AI will transform how we interact Stay up-to-date on the topics you care about. We'll send you an email alert whenever a news article matches your alert term. It's free, and you can add new alerts at any time.

Dutch airline KLM has partnered with DigitalGenius to help incorporate machine learning into its customer service, but humans are in no way getting replaced. Robots remain a common theme in the'who will take your job?' discussions accompanying much of the debate about the future of work. However, very few have looked at the complementary rather than the industrial, revolutionary side of things. Monotonous, mundane tasks have forever come under attack from technology, with machines now operating across most manufacturing lines – though that doesn't mean we should all fear change. Take customer care, for example: it's something that seems menial, but would struggle to fully work on automation.

With growing interest in neural networks and deep learning, individuals and companies are claiming ever-increasing adoption rates of artificial intelligence into their daily workflows and product offerings. Coupled with breakneck speeds in AI-research, the new wave of popularity shows a lot of promise for solving some of the harder problems out there. That said, I feel that this field suffers from a gulf between appreciating these developments and subsequently deploying them to solve "real-world" tasks. A number of frameworks, tutorials and guides have popped up to democratize machine learning, but the steps that they prescribe often don't align with the fuzzier problems that need to be solved. This post is a collection of questions (with some (maybe even incorrect) answers) that are worth thinking about when applying machine learning in production.

Sirajology 23,273 views It Started Out with a Brick - Genetic Algorithm - LD36 - Duration: 3:07. Pete He 1,775 views Nova Science: A new Discovery of the Universe Documentary HD 1080p - Duration: 53:59.

This is the code for Genetic Algorithms by @Sirajology on Youtube. In this demo code we use the MAGIC Gamma Telescope dataset to build a classifer. The classifier will train on the dataset and then be able to classify whether or not some energy is either Gamma Radiation or Hadron Radiation. Instead of guessing and checking the best ML model and hyperparameters to use, we use a genetic programming library called tpot to do that for us by trying out a bunch of them. See this link for an IPython notebook version of this code.

And could software design book covers that could be judged--correctly--by humans? Research in Japan says maybe. 'Designed To Be Judged' At At MIT Technology Review, an article from looks at work being done by Brian Kenji Iwana and Seiichi Uchida at Kyushu University in Japan. According to the report, they've trained a deep neural network to study book covers--to see if the network can identify genre. "Book covers are designed to give readers an idea of the content," after all, the report points out. "Good book covers are designed to be judged.

Want to watch this again later? Report Need to report the video? Report Need to report the video? Need to report the video? This feature is not available right now.

Today we can collect lots and lots of performance data. We build beautiful dashboards and even have fancy query languages to access and transform the data. Still performance data is a secret language only a couple of people understand. The more business becomes digital the more stakeholders are interested in this data including how it relates to business. Some of these people have never used a monitoring tool before.

The new Adobe Voco app is another example of how Digital data is different from the physical world. Digital information has no physical mass and can be edited, recoded, modified. It is this ambivalence feature that is so in contradictory to our everyday experience; we wish we can change our appearance, alter our skills or alter ego but our physical human bodies Can not yet be easily changed. This same property of digital data is what enables cyber security hackers and theft to be done remotely and unseen insider networks and software Or surveillance through web cams or now the growing internet of things of smart home devices that talk, recognise our faces, thumb prints and much more potentially. This is changing our boundaries of privacy and enabling new capabilities that could be for powerful good in areas of medicine to protection from car crashes to virtual reality entertainment.

Smartphone-based virtual reality headsets are great and all, but for the best games and experiences you need a dedicated facehugger tethered to a powerful PC like it's a diver's lifeline. Wireless hardware is one of the inevitable next steps for VR, and a company called TPCAST is already developing a cord-cutting peripheral for the Vive, supported by HTC's VR accelerator program. MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) is making headway in this area too, today releasing research into a wireless system that's both headset-agnostic and could address some unforeseen problems with peripherals like TPCAST's. MIT CSAIL's prototype system, known as MoViR, uses millimeter waves to send data from a transmitter that's hooked up to a computer to the headset's receiver. These high-frequency radio waves are capable of maintaining wireless connections at speeds over 6 Gbps -- enough bandwidth to stream the two, high-definition feeds required for VR -- but the signal doesn't penetrate objects well.