AI Biweekly: 10 Bits from Apr W 2 – Apr W 3

April 3rd – IBM Boosts Oil and Gas Sector Efficiency With AI

IBM announces a four-year program in collaboration with Calgary-based Natural Resources Solutions Center to help oil and gas companies with sustainability and efficiency. IBM’s AI platform Watson will digest raw data, structure it and come up with insights for oil and gas companies.

April 3rd – Amazon Adds to Alexa’s Routine Function

Amazon adds a musical upgrade to Alexa’s “routine” function, which enables users to combine multiple actions such into one command — for example to turn on both the coffee maker and kitchen lights in the morning. Alexa can now play music, podcast or radio shows as part of a routine, and users can also control audio output on the device. The move brings Alexa up to par with Google Assistant, which already integrates radio, podcasts and music.

April 4th – Element AI Announces New Toronto Office

Montreal-based incubator Element AI will open a new office in Toronto where it will do R&D and work with local businesses that want to integrate AI into their operations and the Ontario Provincial Government. Element AI’s July 2017 fundraising round raised a record US$137.5 million. This will be Element AI’s fifth office, and the company is considering opening more in Asia in the near future.

April 6th – California Introduces New Rules for Self-Driving Cars

California’s Public Utilities Commission has introduced a proposal to allow autonomous vehicles without a backup driver to transport passengers on state roads. The California Department of Motor Vehicles had already approved autonomous driving tests without a backup driver starting this month. The new regulations would allow members of the public to ride in such vehicles, and will be voted on next month.

April 10th – Microsoft Collaborates with C3 IoT to Accelerate AI in Enterprise

Microsoft and software company C3 IoT have announced a partnership to accelerate cutting-edge business level AI and IoT application development in areas such as AI predictive maintenance, dynamic inventory optimization, precision healthcare, and CRM. The co-marketing and co-selling deal also includes a co-development strategy on Microsoft’s Azure cloud platform.

April 11th – Qualcomm Introduces New AI Chipset for IoT

Qualcomm announces its own IoT and AI-optimized system on a chip (SoC) platform. The new platform is aimed at computer vision IoT applications such as security cameras, wearable cameras, and smart displays.

April 11th – AirFusion Launches AI Platform For Damage Detection in Wind Power Industry

AirFusion Wind is a cloud-based workflow and AI-based analysis platform that can identify and classify wind turbine asset damage. AirFusion Wind transforms inspection images from drones, ground-based sensors, and other image capture tools into data that can be used to evaluate conditions and reduce costs.

April 11th – NVIDIA Collaborates with Canon Medical Systems to Accelerate Deep Learning in Healthcare

NVIDIA announces a collaboration with Canon Medical Systems to develop research infrastructure to support deep learning technology in the healthcare industry. The partnership will focus on deploying deep learning and big data analytics to support early detection and assisted diagnosis. Canon Medical Systems is the largest medical systems supplier in Japan, and will use the NVIDIA DGX system to process the massive data generated by its platform.

April 13th – Google’s New NLP Tech Plays Word Games

Google’s natural language processing and synthesis research just got fun. The company has uploaded two playful and creative interactive web experiments based on its word-association systems. These are offshoots of a new search option that allows users to directly ask the system questions instead of searching for specific words, titles, or authors, etc. The results are not perfect, but the system provides a useful and flexible new query function.

April 13th – Facebook Uses AI to Predict and Sell User Preferences

The Intercept creates a stir when it publishes apparently confidential internal Facebook documents that describe how AI can build predictive models of user behaviour that Facebook can sell to brands, a process some argue is unethical. The Intercept article argues that although Facebook says these algorithms are used to improve the user experience, in fact they are largely used to make money from advertisers.