What’s happened in AI: May 14th-20th

By | May 23, 2018

Interesting week in AI to say the least. To start, a school in China is now using AI to catch students who aren’t paying attention and/or napping. RIP to those students if they were anything like me in school.

Meanwhile across the pond, the Metropolitan (U.K) police admitted their facial recognition technology has an abysmal 2% accuracy rate. This isn’t the first U.K police force to have trouble with the technology. I wrote about the South Wales (U.K) police force earlier this month and their 8% accuracy rate. Think it’s time the U.K takes a deeper look into this before innocent people suffer the consequences. This is dangerous stuff.

On the autonomous vehicle front, we saw a ton of action this week. Hyundai’s working on their new sensors, MobilEye started testing its vehicles in Jerusalem, Apple now has 55 self-driving cars in California, China is designing a smart-city centered around autonomous vehicles, Roadstar.ai out of China raised over $100mm and Tencent is now testing its vehicles in Shenzen.

Company developments:

Hyundai Mobis Aims To Develop All Autonomous Driving Sensors By 2020 – May. 20, 2018 (Nasdaq)

  • Hyundai Mobis said that it plans to use its proving ground, which opened last June in Seosan, South Chungcheong Province, as a test bed for new technologies, with the goal of becoming a leader in future car technologies. The automotive supplier announced at the proving ground that it plans to develop autonomous driving sensors, which are essential for future vehicles, by 2020, and lead the global market for autonomous driving based on its technology power
  • The company said it will increase its R&D investment cost from the current level of 7% of its parts sales to 10% by 2021. It will also invest 50% of it in ICT, which is closely related to autonomous driving, expanding related R&D workforce and infrastructure and actively promoting technical partnerships with global companies
  • “We are aggressively investing in autonomous driving technologies. We will also increase our R&D workforce for autonomous driving from the current level of 600 to over 1,000 by 2021 and increase the number of our global autonomous test vehicles, from 3 to 20 by next year,” forecasted Yang Seung-wook, Executive VP of ICT R&D Center at Hyundai Mobis

Inside the training camp where Google shares its A.I. secrets with companies Alphabet invested in – May. 19, 2018 (CNBC)

  • CapitalG started the program this spring as a way to tap one of Google’s biggest assets — its leadership in the field of A.I., or artificial intelligence, where machine learning fits in — as it faces stiffer competition from other gigantic investors, like the $100 billion Softbank Vision Fund, that also want a piece of Silicon Valley’s growth
  • The multi-day program offers a multi-day machine-learning crash course adapted from Google’s own internal training program

Baidu COO Qi Lu steps down, AI chief now reports directly to CEO – May. 18, 2018 (VentureBeat)

  • Baidu today announced that COO Qi Lu will step down in July for personal and family reasons. Lu had been brought aboard to help the Chinese search giant become more centrally focused on AI services.
  • Since he joined Baidu, the company has launched a smart speaker, its Duer virtual assistant, and a $1.5 billion fund to grow its Apollo autonomous driving division
  • As part of a consolidation of its various AI products in the wake on Ng’s departure, Haifeng Wang was named general manager of the AI Group and reported directly to Lu. In the absence of a COO, Wang and other senior executives will now report directly to Li, a company spokesperson told VentureBeat. Wang oversees the company’s efforts in areas like machine learning, computer vision, natural language processing, speech technology, knowledge graph, robotics, and augmented reality

Marlabs Launches New AI & Cognitive Computing Platform – May. 18, 2018 (RT Insights)

  • Digital transformation company Marlabs, Inc has announced the availability of their AI and cognitive computing platform, mAdvisor. According to their announcement, the new platform is positioned as a “Data Scientist in a Box,” a simple solution to the problem of talent shortage in the data scientist field. This shortage makes AI adoption using machine learning and cognitive computing a real challenge
  • mAdvisor uses machine learning, natural language generation, machine reasoning and natural language processing to simulate human behavior that is then used to automate tasks like forecasting, data analysis and predictive analytics. The goal is to improve decision making via faster and more accurate insights

Intel/MobileEye self-driving cars begin testing in Jerusalem – May. 17, 2018 (The Drive)

  • Mobileye is based in Israel, but the company and parent Intel also want to demonstrate that self-driving cars can work in all driving conditions, CEO Amnon Shashua said in a press release. Jerusalem’s aggressive drivers, poorly-marked streets, and complicated merges will require autonomous cars to develop a new set of skills, he said
  • While most self-driving cars rely on a combination of cameras, radar, and lidar, cars deployed in Jerusalem will use only cameras initially. Cars will use 12 cameras, including eight dedicated to long-range sensing, and four for parking. Intel and Mobileye want to prove that a self-driving car can operate with cameras alone, then adding radar and lidar at a later stage will create “true redundancy.” It probably doesn’t hurt that cameras are Mobileye’s specialty
  • Intel and Mobileye hope to put SAE Level 4 and Level 5 autonomous-driving systems on the market by 2021. BMW and Fiat Chrysler Automobiles have already partnered with the two tech companies on research, so they will likely be among the first customers

The World’s Dominant Crypto-Mining Company Wants to Own AI – May. 17, 2018 (Bloomberg)

  • During a rare series of interviews with Bloomberg Businessweek at Bitmain’s headquarters, Wu and his executives say AI chips are a natural adjacency for their privately held, extremely secretive company. “Artificial intelligence requires lots of computations,” says Wu. That makes it like Bitcoin mining, which is best done by a custom chip known as an application-specific integrated circuit, or ASIC
  • On the strength of such chips, Bitmain has quietly become a global power. A February report by Bernstein Research estimated that the company might be generating profits of more than $3 billion per year, about the same as its much larger rival, Nvidia Corp.
  • The report also suggested that Bitmain is probably one of the five biggest customers of the most advanced fabrication process offered by the Taiwanese chip giant Taiwan Semiconductor Manufacturing Co., which also counts among its clients Apple Inc. and Qualcomm Technologies Inc. Wu declined to comment on Bitmain’s profits, but it seems a safe bet that, given the company’s revenue figure, things aren’t quite as rosy as Bernstein’s estimate. TSMC declined to comment

Healthcare Sales Executive, Daniel Gauvreau, Joins Inspirata Canada as Vice President, Sales and Marketing for AIM, Inc – May. 17, 2018 (Global Newswire)

  • Cancer informatics and digital pathology workflow solution provider Inspirata®, Inc. announced today that healthcare sales executive, Daniel (Dan) Gauvreau, has joined the company’s Canadian subsidiary, Artificial Intelligence in Medicine, Inc. (AIM), as Vice President, Sales and Marketing, effective May 7, 2018
  • Based in the company’s Canadian headquarters in Toronto, Dan will be responsible for Canadian-based sales of AIM’s Artificial Intelligence (AI) and Natural Language Processing (NLP) tools that extract cancer-related data from clinical documents, such as pathology reports, molecular testing reports, treatment plans and clinician’s notes. He also will gradually assume responsibility for sales and customer support for Inspirata’s digital pathology workflow solution in Canada

LinkedIn launches AI Academy to bolster internal AI skills – May. 16, 2018 (ZDNet)

  • LinkedIn is launching an internal program called AI Academy that aims to teach employees across the company how to incorporate artificial intelligence tools into their daily work
  • LinkedIn’s motivation for launching the AI Academy traces back to the Silicon Valley talent wars. Tech companies are scrambling to attract top talent in the areas of AI and machine learning, and often face difficulties scaling AI implementations because of a lack of AI skills

China car owners use AI-Powered app to file damage claims – May. 15, 2018 (Alizila)

  • Assessing damage caused to their rides has just gotten a lot easier for car owners in China, with the rollout of a video-based, artificial-intelligence app from Ant Financial
  • The Alibaba affiliate last week launched version 2.0 of its Dingsunbao (Loss Assessment Master) app, giving drivers the same power in their hands to provide detailed car-damage information to insurers and claim vehicle insurance in real time as Ant Financial gave professional insurance adjusters just under a year ago
  • The first version of Dingsunbao has already helped insurers, including China Taiping, China Continent Insurance, Sunshine Insurance Group and AXA Tianping process claims tens of millions of times at a rate of speed much faster than human adjusters alone could handle

Linguamatics I2E AMP 2.0 Automates Critical Insights Discovery in Real-Time with NLP – May. 15, 2018 (SYS)

  • Linguamatics, the leading natural language processing (NLP) text analytics provider, today announced the latest release of its I2E AMP platform to automate the discovery of critical insights from text using NLP
  • The I2E Asynchronous Messaging Pipeline (AMP) platform delivers high-throughput, fault tolerant workflow management for real-time document and record processing, addressing the NLP text-mining and ETL (extract transform load) requirements for healthcare and life science organizations of all sizes by allowing users to plug I2E into enterprise workflows
  • I2E AMP 2.0 includes enhanced functionality to speed overall throughput and performance, sophisticated pre- and post-processing capabilities, a Web GUI to simplify the set-up of initial workflows, and new AMP Agents for smarter load balancing, easier deployment, and optimized I2E management

Dropbox eyes security, machine learning technologies in Israel – May. 15, 2018 (Reuters)

  • U.S. file-sharing and storage company Dropbox (DBX.O) plans to expand its development center in Tel Aviv by hiring more staff and looking out for potential partnerships and acquisition opportunities in security and machine-learning
  • Israel’s education system and military service have created a pool of talented people working in cybersecurity and “that is something we deeply need to take advantage of”, Quentin Clark, who leads engineering, product and design at Dropbox, told reporters on Tuesday during a visit to Israel

Automated Insights Continues Rapid Growth, Software Veteran Kevin Fitzgerald Joining Team as Chief Revenue Officer – May. 15, 2018 (PR Newswire)

  • Automated Insights (Ai), the world’s leading natural language generation (NLG) company, today announced the hiring of Kevin Fitzgerald as Chief Revenue Officer. Fitzgerald will be replacing Chris Neal, who joined Automated Insights in May of 2017 to define and grow Ai’s enterprise sales organization
  • With that goal accomplished, Fitzgerald will build on Neal’s success to rapidly scale the sales organization as Ai bolsters its industry-leading status and continues its growth in the world of business intelligence. Chris will continue with Ai on a full-time basis for the remainder of the first half of 2018 and then transition to a consulting role thereafter

Shenzhen gives Tencent license to road test autonomous vehicles – May. 14, 2018 (TechNode)

  • Tencent has obtained a license for road tests of intelligent connected vehicles from Shenzhen municipal government, our sister site has reported (in Chinese). Tencent said that the transportation bureau issued only one license this time. The company will begin testing autonomous vehicles on public roads after the government designates the testing area
  • The Tencent Autonomous Vehicle Laboratory made its debut last November. The laboratory operates under Tencent’s Mobile Internet Group (MIG) along with other businesses such as connected car services, location services, and Tencent Maps. The company plans to further integrate other resources—including cloud, security, big data, and AI—into the autonomous vehicle platform

Tencent unveils China’s first open source Go AI, made on spare WeChat server processing power – May. 14, 2018 (TechNode)

  • Tencent has unveiled its Go-playing AI program, “PhoenixGo,” on Github, making its source code and training model available on the open source platform (in Chinese). The program, developed by WeChat’s translation unit, is the first open source Go AI project in China
  • A team of WeChat engineers focusing on the development of machine translation decided to use the spare processing power of WeChat’s servers to train the PhoenixGo program, which implemented AlphaGo Zero, the latest version of DeepMind’s champion-defeating AlphaGo, and increased its training efficiency. PhoenixGo’s source code and training model can run on a single GPU chip and perform at the same level as a professional Go play.er

Apple’s self-driving car fleet grows to 55 in California – May. 14, 2018 (TechCrunch)

  • Apple now has 55 self-driving cars registered with the California Department of Motor Vehicles, compared to 27 earlier this year and just three last year. That means Apple has the second largest fleet of self-driving cars in California, Mac Reports first pointed out
  • Apple now has more cars registered than Waymo, which has 51, according to the California DMV. General Motor’s Cruise, however, leads the pack with 104 vehicles. In total, the DMV has provided to 53 companies permits for self-driving cars that include safety drivers, resulting in a total of 409 vehicles and 1,573 safety drivers

Google’s DeepMind is using neural nets to explore dopamine’s role in learning – May. 14, 2018 (VentureBeat)

  • Researchers at DeepMind in London modeled human physiology using a recurrent neural network, a type of neural network that’s able to internalize past actions and observations and draw from those experiences during training
  • The reward prediction error of the network — the signal that mathematically optimizes the algorithm over time through trial and error — stood in for dopamine, the chemical in the brain that affects emotions, movements, and sensations of pain and pleasure, which is thought to play a key role in the learning process
  • “Leveraging insights from AI which can be applied to explain findings in neuroscience and psychology highlights the value each field can offer the other,” the DeepMind team wrote. “Going forward, we anticipate that much benefit can be gained in the reverse direction, by taking guidance from specific organization of brain circuits in designing new models for learning in reinforcement learning agents.”

GraphPath plans to combine Knowledge Graphs with the blockchain – May. 14, 2018 (TechCrunch)

  • Knowledge Graphs are essentially a way to organize data in an endless network of facts and relationships. It’s similar to Facebook’s ‘social graph’ (which contains highly descriptive data about how people are connected) but instead is about how knowledge is connected. In theory, you could use blockchain to create a knowledge graph, and in fact, that’s what the company claims to have now done. Google’s Knowledge Graph powers its search functions, and according to the company contains over 70 billion facts
  • Today GraphPath is announcing GraphOS, which is described as “the world’s first decentralized operating system for Knowledge Graphs on the blockchain, that will be enabled by the GraphOS Protocol,” according to Demian Bellumio, GraphPath’s CEO
  • GraphOS will allow for a network of Knowledge Graphs to be safely and securely be interconnected via blockchain technology. Through this network, a wide array of participants, such as Knowledge Graph owners, data providers, knowledge experts, and data scientists will be able to transact with each other. GraphPath calls this ecosystem, a Graph Operating System, as it will manage all the resources necessary to run the Network, thereby the name GraphOS


Ericsson and SoftBank deploy machine learning radio network – May. 17, 2018 (ZDnet)

  • The machine intelligence-powered service reduces telco operational expenditure and improves network performance by grouping cells and analysing cell overlap, signal strength, and receive diversity to make decisions on carrier aggregation between cells, Ericsson said
  • “SoftBank was able to automate the process for radio access network design with Ericsson’s service. Big data analytics was applied to a cluster of 2,000 radio cells, and data was analysed for the optimal configuration.”
  • Head of Managed Services Peter Laurin said Ericsson is investing heavily in machine learning technology for the telco industry, citing strong demand across carriers for automated networks. To support this, Ericsson is running an Artificial Intelligence Accelerator Lab in Japan and Sweden for its Network Design and Optimization teams to develop use cases

4C and DigitalGenius Launch AI Partnership – May. 14, 2018 (PR Newswire)

  • 4C, EMEA’s leading independent Salesforce Platinum Partner and DigitalGenius, the Artificial Intelligence (AI) platform for customer service, today announced a new partnership that will allow both businesses to meet the growing demand for AI services, while bringing practical applications of data science and AI into the customer service operations of many leading companies across EMEA
  • AI and Data Science have always been a core practice for 4C. This new strategic partnership with DigitalGenius will strengthen this and allow 4C to advance and further develop their machine learning and AI offering while bringing their customers to the forefront of this new technological frontier
  • Currently, 4C has a team of more than 15 AI specialists working on a variety of projects, including the recent optimisation of Dutch insurance company, NN’s customer service centre, with the implementation of 4C’s AI model, TellMi. DigitalGenius is recognized as a leader in the AI space and is already powering some of the world’s most innovative customer service organizations with practical applications of AI


Microsoft acquires conversational AI startup Semantic Machines – May. 20, 2018 (VentureBeat)

  • Microsoft today announced that it has acquired Semantic Machines to bolster its conversational AI offerings — like Cortana, the Azure Bot Service, and Microsoft Cognitive Services. Semantic Machines works in areas like speech synthesis, deep learning, and natural language processing
  • Semantic Machines describes itself as a company bent on creating conversational AI that enables machines “to communicate, collaborate, understand our goals, and accomplish tasks.” It could help Microsoft compete with conversational computing initiatives from Amazon’s Alexa, Apple’s Siri, Google’s Assistant, and Samsung’s Bixby
  • In addition, Semantic Machines has assembled a cadre of experts in the conversational AI arena, like Larry Gillick, former chief scientist for Siri at Apple, and well-known researchers like UC Berkeley professor Dan Klein and Stanford University professor Percy Liang. “With the acquisition of Semantic Machines, we will establish a conversational AI center of excellence in Berkeley to push forward the boundaries of what is possible in language interfaces,” Microsoft AI and Research Group CTO David Ku said in a blog post

Dialpad acquires TalkIQ for AI that transcribes conferences calls – May. 16, 2018 (VentureBeat)

  • Dialpad today announced it has acquired TalkIQ, a company that offers conversational AI services for meetings, including transcriptions and the automation of identifying action items in meeting transcripts
  • TalkIQ’s platform and natural language processing and speech recognition services are now known as VoiceAI. They will be incorporated into all Dialpad products, CEO Craig Walker told VentureBeat in an interview, including Uberconference, Dialpad phone replacement, and its offerings for call centers


Chinese self-driving tech startup Roadstar.ai raises industry record funding of $128m – May. 16, 2018 (Deal Street Asia)

  • Roadstar.ai, a Shenzhen-based AI-powered self-driving technology startup founded by former Google and Tesla engineers, announced that it has raised $128 million in Series A funding round, the biggest-ever investment made into a Chinese startup in the country’s autonomous driving industry
  • In a statement posted on its website, Roadstar.ai said the latest funding round was co-led by venture capital firm Shenzhen Capital Group and Wu Capital, a family office in Mainland China. CMB International Capital, Vision Capital and early investors Glory Ventures, Shenzhen Green Pine Capital, Ventech, and Yunqi Partners also participated in the round
  • The startup has developed two key technologies – HeteroSync and DeepFusion. It claims that HeteroSync provides accurate time and spatial synchronisation, real-time update, and robust data-based features. DeepFusion, meanwhile, produces robust, efficient, and safe autonomous driving solution

China’s iFlytek To Raise $567M To Expand Artificial Intelligence Investment – May. 15, 2018 (China Money Network)

  • Chinese artificial intelligence company iFlytek Co Ltd., best known for its voice-recognition technology, plans to raise up to RMB3.6 billion (US$567 million) through private placement to expand its artificial intelligent investments, according to its security filing
  • The proceeds of the fund will be used for the development of the new generation of cognitive technology, AI speech open platform, smart robots, sales and service upgrade, as well as improving liquidity, the company said
  • The biggest investment of RMB1.18 billion (US$186 million) will go to the development of a smart speech open platform . The project will take three years to complete. It is expected the generate revenue of RMB3.09 billion (US$487 million) and net profit of RMB532 million (US$83.8 million) per year once completed

BrainQ raises $5.3M to treat neurological disorders with the help of AI – May. 15, 2018 (TechCrunch)

  • BrainQ, an Israel-based startup that aims to help stroke victims and those with spinal cord injuries treat their injuries with the help of a personalized electromagnetic treatment protocol, today announced that it has raised a $5.3 million funding round on top of the $3.5 million the company previously raised
  • The company’s investors include Qure Ventures, crowdfunding platform OurCrowd.com, Norma Investments, IT-Farm and a number of angel investors, including Valtech Cardio founder and CEO Amir Gross
  • When we last talked to BrainQ earlier this year, the team was working on two human clinical trials for stroke patients in Israel. At that time, the company had closed its first funding round and had also recently started to work with Google’s Launchpad Accelerator, too

Beautiful.AI fetches $11mm Series B – May. 15, 2018 (PE Hub)

  • Beautiful.AI, which leverages artificial intelligence to automate the visual design process, has raised $11 million in Series B funding. Trinity Ventures led the round with participation from return backers Shasta Ventures and First Round Capital. In addition to the funding, Karan Mehandru, a general partner at Trinity Ventures, will join Beautiful.AI’s board of directors

OspreyData’s AI-Based Optimization Platform for Oil & Gas Raises $5M in Series A Funding Led by Houston Ventures – May. 15, 2018 (VentureBeat)

  • OspreyData announced a $5 million Series A growth financing led by Houston Ventures with participation from existing shareholders. “This investment will help support our technology leadership position and the further adoption of our unique Expert-Augmented Machine Learning methodology and platform. We will continue to grow the sales and services teams to meet the needs of our global Oil & Gas customers,” commented OspreyData CEO John Burke
  • Fred Lummis, Junior Partner at Houston Ventures, said OspreyData was a perfect fit for their fourth fund that targets inefficiencies in Oil & Gas operations. He states, “OspreyData’s hardware-agnostic analytics platform has the ability to deliver massive ROI by leveraging existing production asset data streams to transition operators to an active monitoring program that reduces asset downtime and optimizes well performance

Artificial intelligence startup Motorleaf closes a $2.85 million seed round to accelerate AI in greenhouse operations – May. 15, 2018 (Tech Startups)

  • Motorleaf, announced it has raised $2.85 million USD to further develop its real-time, agronomic solutions for the fast-growing controlled-environment agriculture sector
  • The investment, closed in two rounds with the final $2 million USD committed this May, is led by top ag- and food-tech acceleration fund Radicle Growth with support from Desjardins Capital, Real Ventures, Fluxunit (Osram Ventures), BDC Capital and 500 Startups Canada
  • MotorLeaf has been built primarily to help growers produce more, and a better quality product; by automatically adjusting to the needs of your crop. MotorLeaf builds separate components, each of which has a specific set of functions. They can operate separately, or if you add onto the MotorLeaf HEART – operate as a larger whole. With automated adjustments to your crop, and notifications on when there may be a problem for you to look into – you decrease your loss, saving / making you more money


New PCG Research on the Automotive Industry Finds That Conversica’s AI Solutions Are ‘Bursting With Exponential Opportunity for Dealers’ – May. 17, 2018 (PR Web)

  • Conversica, the leader in conversational AI for business, today announced a new report by PCG Research which concluded that Conversica’s AI sales and service assistants improve customer satisfaction, generate greater revenue opportunities, influence higher service department traffic, and ultimately perform as a critical customer retention tool for automotive dealerships of all sizes
  • “When AI and machine learning are put into the context of the auto industry, they are bursting with exponential opportunity for dealers,” said Brian Pasch, CEO of PCG Research. “Conversica has been on the leading edge of harnessing that power, offering dealers sophisticated yet easy-to-implement tools that significantly impact sales and service revenue and customer retention.”

Machine learning algorithm is claimed to predict which students will drop out – May. 14, 2018 (Stuff)

  • A machine learning algorithm which predicts the likelihood a student will drop out of university is being trialled in New Zealand and Australia
  • Using 15 years’ of historical student data, Jade Software data scientists built the tool in just four weeks as a research project. The team applied various machine learning algorithms to the data, which included information such as students’ grade point average, age, distance of residence from campus, enrolment history, ethnicity, and how they pay for their studies
  • “[The university] will get a list of students ranked from most likely to drop out to least likely, and they can work from the top, essentially to reach out,” he said. “It’s about informing universities and giving them a report or on-demand red flag saying this person has a 60 per cent or 80 per cent chance of dropping out, and then allowing the universities to intervene.”

Government / Policy

China intends for self-driving cars to propel smart megacity – May. 20, 2018 (Nikkei Asia Review)

  • China envisions self-driving technology for all private cars in a cutting-edge metropolis to be completed near Beijing by 2035, taking aim at U.S. supremacy in autonomous vehicles and pitting authority-driven progress against free market innovation
  • Designing the new megacity around self-driving automobiles eliminates problems associated with adapting existing infrastructure to cope with a mix of regular and automated vehicles — as well as pedestrians. China hopes to use Xiongan as a model for future cities worldwide

Trudeau touts Canada’s AI credentials at MIT tech gathering – May. 18, 2018 (Tampa Bay Times)

  • “We’ve been investing massively in AI,” Trudeau told a conference of tech entrepreneurs Friday at the Massachusetts Institute of Technology, before name-dropping several of the Canadian academics he said provided the underpinnings of the technology during an “AI winter” when most weren’t paying it much notice
  • He said Friday that Canada is making a “deliberate choice,” but not an easy choice, to embrace change at a time when new technology is disrupting workplaces and leading to anxiety and fear about the future. He said leaders also have a responsibility to shape the rules and principles to guide the development of artificial intelligence
  • His visit headlined MIT’s Solve initiative, which connects innovators with corporate, government and academic resources to tackle world problems. He also spent time visiting a technology lab, chatting with engineers and asking pointed questions about hydroponics, neurology, robots and kids’ coding programs

Tech-savvy Chinese high school catches napping students using AI cameras – May. 17, 2018 (Straits Times)

  • The Hangzhou No. 11 Middle School has installed a “smart classroom behavior management system”, which captures students’expressions and movements, analyzing them with big data to make sure they are paying attention, media reported on Thursday (May 17)
  • The system will be able to tell if students are reading or listening – or napping at their desks. It can detect expressions like happiness, repulsion, fear, anger and befuddlement. Students will get a real-time attentiveness score, which will be shown to their teacher on a screen, media said
  • The system is devised by Hikvision Digital Technology , one of the world’s biggest suppliers of security cameras and which is developing its own AI technology

Tianjin city in China eyes US$16 billion fund for AI work, dwarfing EU’s plan to spend US$1.78 billion – May. 16, 2018 (South China Morning Post)

  • Tianjin, a northeastern port city outside Beijing, said it will establish a 100 billion yuan((US$15.7 billion))fund to speed up development of new generation AI technologies, according to a document posted on the city government’s official website on Wednesday
  • The money will be raised from financial organisations in China and abroad as well as from private enterprises and other market players, it said. The AI industrial fund is part of Tianjin government’s grand plan to “speed up the development of the smart technology industry”, which includes a 10 billion yuan smart manufacturing fund financed by the municipal government and a string of talent incentives – up to 2 million yuan cash reward per person – to attract high-end tech talent

New York City announces task force to find biases in algorithms – May. 16, 2018 (VentureBeat)

  • New York City mayor Bill De Blasio today announced a task force to examine how city agencies use algorithms to make decisions. The mayor’s office called it the “first of its kind in the U.S.” and said the task force aims to produce a report in December 2019
  • Specifically, the new task force will examine machine learning algorithms used by the New York City Police Department, the Department of Transportation, the Mayor’s Office of Criminal Justice, the Administration for Children’s Services, the Department of Education, and the Department of Social Services. The task force will develop a process to determine whether the automated systems are “fair,” “equitable,” and “accountable.” The report will also identify ways the algorithms’ decision-making can be made more transparent
  • The creation of the task force is mandated by a City Council law passed in 2017, as part of a broader effort to study how AI systems discriminate against people based on race, gender, sexual orientation, and citizenship status. Council member James Vacca told ProPublica he was inspired to sponsor the bill by investigations into a racially biased algorithm designed to classify DNA samples

Met Police’s facial-recognition software wrong for 98% of matches, report says – May. 14, 2018 (Alphr)

  • Facial-recognition tech used by the Metropolitan Police is wildly inaccurate, flagging up false positives in 98% of the cases
  • Similarly, the facial-recognition software being trialled by the South Wales Police is also highly error-prone. Out of the 15 deployments made by the police force since June 2017, there have been 2,400 false positives. Only 234 (10%) of the software’s matches have been accurate
  • With results such as these, it’ll be difficult to gain public acceptance of a technology that is, currently, highly unreliable. Silkie Carlo, director of the Big Brother Watch pressure group, believes that the technology isn’t fit for purpose, and “police must immediately stop using real-time facial recognition if they are to stop misidentifying thousands of innocent citizens as criminals.”


Transform, the AI event for growth (August 21-22) – May. 17, 2018 (VentureBeat)

  • This flagship event, previously called MobileBeat, will focus on transformative technologies like AI and data analytics that are revolutionizing the tech marketing ecosystem and delivering results in customer acquisition, retention, and engagement