October 18th – Google’s AlphaGo Zero Masters the Game of Go with No Human Knowledge
Google’s DeepMind publishes a paper in Nature introducing the latest evolution of its AI-powered Go program. “AlphaGo Zero” learns in self-play games, with no human knowledge required. The program crushed previous “AlphaGo” versions (including the one that beat world-best human Ke Jie) with a record of 100 wins and zero losses, stimulating discussion in the Go and AI communities. The new AlphaGo Zero uses just four Tensor Processing Units (TPU), and was trained with only five million self-play games over a few weeks.
October 18th – Apple and GE Partner to Bring Predix Industrial Apps to iPhone and iPad
Apple and GE reveal their partnership to develop robust industrial applications to bring more data to GE’s IoT platform, Predix. The platform will release more SDKs for iOS, which developers can generate industrial apps on the iPhone and iPad. Industrial operators can efficiently access the data and gain insight into the status of their equipment. Industrial data can be visualized and analyzed under Apple’s ecosystem.
October 24th – Delphi Buys Autonomous Driving Technology Firm NuTonomy for US450 Million
Delphi Automotive (Delphi), one of the world’s largest automotive parts manufacturers, announces the acquisition of MIT spin-off self-driving car company nuTonomy for US$450 million. An earnout of $50 million will be retained until the company proves its profitability. UK-based Delphi entered the autonomous driving race in 2011 by investing heavily in sensors.
October 25th – Philips and Huawei Initiate Cloud AI Healthcare Project
China’s mobile giant partners up with Dutch electronics company Philips. The main goal of this collaboration will be to deploy cloud-based AI healthcare solutions targeting small urban centers in China. After implementing Cloud AI, doctors can work more efficiently in terms of information processing and sharing.
October 25th – LinkedIn is Training Its Engineers with Artificial Intelligence
LinkedIn opens an academy to train its engineers in a fundamental knowledge of AI. LinkedIn’s head of AI Deepak Agarwal says that “Everyone wants to have AI as a component of their product.” Six engineers have already passed through the training. The eventual goal is to implement AI into LinkedIn’s existing product.
October 26th – Dell Technologies Puts US$1 Billion into the Internet of Things
Dell confirms its US$1 billion investment in IoT product R&D over the next three years. The investment will not only enhance solution development but will also aim to build an end-to-end ecosystem.
October 26th – Walmart Installs Shelf-Scanning Robots
Inventory inaccuracy has been a chronic challenge for retailers, more so as stores have been gradually performing an order fulfillment role as retailers move online. Walmart’s new robots scan aisles to check for missing or misplaced stock items, incorrect prices and mislabeling, and send restocking or other orders to employees. While robots are 50% more productive than humans in identifying missing and incorrect items on the shelves, Walmart says they will not take retail jobs away from their employees.
October 27th – MIT Media Lab’s AI Can Now Write Horror Stories
MIT Media Lab’s Scalable Cooperation researchers develop an AI to create horror stories in collaboration with humans. The bot was trained on 140,000 horror stories taken from Reddit. It works by writing an opening line to a horror story and posting it on Twitter. Followers can respond with next line of the story, which the bot will learn from and then respond to with a third line etc. The stories have been successful and are being collected on www.shelley.ai. The bot is called Shelley, after Frankenstein author Mary Shelley, and has appeared just in time for Halloween.
October 27th – Microsoft Deploys Healthcare NeXT with UPMC
Microsoft announces its latest initiative in healthcare technology. NeXT is a cloud-based artificial intelligence project with the goal of developing digital tools and transforming research. NeXT’s first partnership is with the UPMC (University of Pittsburgh Medical Center), in a collaboration aiming to improve clinician productivity. Other projects were also announced, including personalized healthcare and chatbot services.
October 22nd – JPMorgan Utilizes AI to Help Traders Predict the Market
JP Morgan is bringing AI to the front of fixed-income salespeople and traders, aiming to provide a picture of the whole trading floor in real time and help them anticipate market moves. AI and machine learning are being introduced to improve salespeople’s performance and not to replace them, the company says.
Credit: Synced Global Industry Analyst Team | Editorial & Infograph: Meghan Han
0 comments on “AI Biweekly: 10 Bits from Oct (Pt 2)”