Downpour SGD (with Adagrad adaptive learning rate) outperforms Downpour SGD (with fixed learning rate) and Sandblaster L-BFGS. Moreover, Adagrad can be easily implemented locally within each parameter shard. It is surprising that Adagrad works so well with Downpour SGD as it was not originally designed to be used with asynchronous SGD. The paper conjectures that "Adagrad automatically stabilizes volatile parameters in the face of the flurry of asynchronous updates, and naturally adjusts learning rates to the demands of different layers in the deep network."

AI is getting bigger and smarter in Seattle. The Allen Institute for Artificial Intelligence, a sister company to Paul Allen's Institute for Brain Science, plans to hire 25 people in the next year as it prepares to take its Aristo technology to the eighth grade, moving on from teaching it fourth-grade science.

France has many hidden gems, and Facebook is well-aware of that. The company is building a new artificial intelligence research team in Paris in order to work on ambitious futuristic projects. The new team will work closely with existing Facebook AI Research (FAIR) teams in Menlo Park and New York on image recognition, natural language processing, speech recognition, machine learning, live translating tools and more. There are a couple of reasons behind this choice. First, France has a lot of incredibly talented researchers -- it is one of the leading countries when it comes to math, physics and robotics research.

Robot workers ready to take your call? No one expected a computer to beat a human at the ancient Korean board game Go for another few years at least. So when Google's "AlphaGo" artificial intelligence won against champion player Lee Sedol last month, there were ripples of shock and awe. A far more complex game than chess, this was a "holy grail" moment for machine learning, an important milestone in history. South Korean professional Go player Lee Sedol, right, puts the first stone against Google's artificial intelligence program, AlphaGo, during the Google DeepMind Challenge Match in Seoul.

I tweet about AI, science fiction, etc. Spoiler alert!

Elasticsearch is an open source search engine based on Lucene. Its being used by leaders in the market like Wikipedia, Linkedin, ebay etc. To install, just unzip the downloaded file and run bin/elasticsearch.bat To install on unix systems you have to run bin/elasticsearch. An index is like a'database' in a relational database.

McCarthy was a true "progressive" in that he appreciated the rapid and dramatic improvements in human living standards brought about by innovation. It was from McCarthy's website that I first learned of Thomas Babington Macaulay's remarks, in the Edinburgh Review, that I often quote -- "We cannot absolutely prove that those are in error who tell us that society has reached a turning point, that we have seen our best days. But so said all before us, and with just as much apparent reason ... On what principle is it that, when we see nothing but improvement behind us, we are to expect nothing but deterioration before us".

The creators of a lifelike "sexy robot" were left red-faced when a technical glitch captured on CNBC led the android to claim it will "destroy humans." David Hanson, founder of Hanson Robotics, was demonstrating "Sophia" at the South by Southwest (SXSW) technology show in Texas when the slip-up occured . After jokingly asking "do you want to destroy humans?...Please say'no'," Hanson was answered by the unblinking machine. It's a far cry from the actual reason Hanson Robotics is developing the lifelike Androids. According to the company, the'bots will serve in theme parks and care facilities to assist people in both customer service and health treatment.

Envisaging a technological future where computers can learn human language and have conversations with people, Microsoft's India-born CEO Satya Nadella has said companies need to build "respectful" technology which gets the best of humanity and not the worst. "We want to take the power of human language and apply it more pervasively to all of the computing interface and interactions," Mr. Nadella said yesterday in his keynote address to thousands of developers at Microsoft's annual Build 2016 conference here. "To do that you have to infuse (intelligence) into the computers around us, you have to bring forth these technologies of artificial intelligence in machine learning so that we can teach computers to learn the human language, have conversational understanding, teach them about the broad contexts of personal preferences and knowledge so that they can help you with your everyday task," Mr. Nadella, 48, said. He envisaged a future where the machine is not against man but works with humans to offer seamless interface and experience in day-to-day learning. "All technology that we built has to be more inclusive and respectful. We want to build technology that gets the best of humanity and not the worst," Mr. Nadella said.

Some of the more interesting and universally accessible developments in technology over the past decade (to me anyway) have been within the artificial intelligence world. Unlike many other industry advancements, stories of high-profile AI successes or failures make suitable discussion fodder at family dinner tables. The "Fork Me On GitHub" types appreciate the complexity of the required engineering efforts, while the "Hey Person Who Computers -- Can You Set My VCR Clock?" crowd can just think "whoa" without concern for what is happening under the hood. One somewhat recent and well-publicized AI event was the Jeopardy-playing Watson supercomputer built by IBM. Those in the technology community understood the difficulties of natural language processing (all while being required to answer in the form of a question) while much of the world just wanted to see a computer built by anonymous geeks humiliate more smug geeks on live TV.