This week was relatively light compared to others, but nonetheless, there were some important developments. In particular, Microsoft revealed another failed chatbot. This time it’s so politically correct that it’s cringeworthy. Speaking of failed AI technology, Rep. Jerrold Nadler and Sens. Ron Wyden, Chris Coons, Ed Markey and Cory Booker are pushing for an investigation into the use of facial recognition technology in law enforcement given the recent cases of false positives. Like many AI technologies, a healthy amount of proper regulation would be good.
Company developments
Driveless Cars With Russian Soft May Appear on Chinese Roads in 2019 – AI Firm – August 3, 2018 (Sputnik News)
- Driverless cars with Russian software may appear on the Chinese market in 2019, Chinese firms do not have such technologies yet, Olga Uskova, head of the Cognitive Technologies company specializing in advanced driver assistance systems (ADAS) based on artificial intelligence (AI), told Sputnik
- According to Uskova, the company is currently engaged in talks with two large Chinese carmakers interested in fourth-stage ADAS, a sort of “brain” that ensures that driverless vehicles can share the road with other cars
- Cognitive Technologies is planning to open a sales office in China and enter the country’s market with agricultural robotics, AI software that can be installed on any type of agricultural machinery. Uskova said that the company’s strategy advantageously coincided with Beijing’s plans to equip as many farms as possible with robots
Xiaomi launches AI-powered feature phones QIN 1, QIN 1s at CNY 199 and CNY 299 – August 3, 2018 (First Post)
- While the Reliance JioPhone has emerged as a clear leader in the Indian mobile handset market with a 27 percent share in the second quarter of 2018, a new feature phone that just launched in China could give it some competition, if it launches in India
- Xiaomi just launched an AI-powered feature phone series called the Qin AI in China. Two phones have been announced as part of the series so far, these include the Qin 1 and Qin 1s and they are not just any regular feature phones. They come with real-time translation support, with services up to 17 international languages. The AI answers all the users questions too
- Neither of them has any camera on the rear or the front, which the JioPhone does. This could mean that even if the Qin AI enters the Indian market, it might not beat Jio, even though it has all those AI capabilities
Google’s China plans include AI-curated news app as Googlers fiercely debate the issue – August 2, 2018 (9 to 5 Google)
- Multiple publications have since confirmed that Google is planning to launch a Search app that would comply with China’s censorship laws. A new report today details that Google is also working on a news-aggregation app that would similarly self-censor content in a sign of grander ambitions for the country. Meanwhile, since details first came to light yesterday, Google employees have been fiercely arguing the issue
- According to The Information this morning, this news app is a part of the Dragonfly initiative within Google to return to China. In fact, it would launch ahead of the Search service to demonstrate to Chinese officials that the company is willing to follow local rules
- The news app would work in a similar fashion to block unsanctioned content. In terms of functionality, the app is quite standard and sounds quite similar to the latest Google News. Today’s report describes how artificial intelligence would be used to personalize content for users in lieu of human editors
Tesla is building its own AI chips for self-driving cars – August 1, 2018 (TechCrunch)
- “We’ve been in semi-stealth mode on this basically for the last 2-3 years,” said Elon Musk on an earnings call today. “I think it’s probably time to let the cat out of the bag…”
- The cat in question: the Tesla computer. Otherwise known as “Hardware 3,” it’s a Tesla-built piece of hardware meant to be swapped into the Model S, X and 3 to do all the number crunching required to advance those cars’ self-driving capabilities
- “The key,” adds Elon “is to be able to run the neural network at a fundamental, bare metal level. You have to do these calculations in the circuit itself, not in some sort of emulation mode, which is how a GPU or CPU would operate. You want to do a massive amount of [calculations] with the memory right there.”
KanKan AI Chosen by AIO to Support “All-in-One” Solution for Analyzing and Forecasting Motion Picture Performance Across China Film Market – August 1, 2018 (PR Newswire)
- Remark Holdings, Inc. (NASDAQ: MARK), a diversified global technology company with leading artificial intelligence (AI) solutions and digital media properties, including Vegas.com, today announced that it has been chosen by Allinone (AIO), a leading big data marketing and consulting company serving China’s film market, to support the industry’s first and only full-process system covering data analysis, forecasting and consulting on film performance from pre-release to post-release across China’s entertainment market
- AIO, whose clients include Warner Media and Paramount Pictures, has been authorized by China’s State Administration of Radio, Film and Television (SARFT) to create a data analysis product using the root level data of China’s film market. Remark has partnered with AIO to utilize its KanKan AI technology to support the design and build-out of the data processing and analytics system. As part of the partnership, Remark has also invested $1 million USD ($6.6 million RMB) in the venture and is providing its technology services. The “All-in-One” solution is being developed to address the highly inefficient film forecasting environment in China, which has resulted in an elevated level of risk for film producers, investors, advertiser and marketers. According to SARFT, China’s film market is expected to be the world’s largest by 2020
Introducing the New Wordsmith Extensions for Tableau by Automated Insights – July 31, 2018 (PR Web)
- Automated Insights, the world’s leading natural language generation (NLG) company, today announced a new integration with Tableau built upon the newly released dashboard extensions feature included in Tableau 2018.2. The Wordsmith Extension enables Tableau customers to interact with their data and communicate insights by pairing data visualizations with written analytics
- “Enabling our customers to leverage the power of Wordsmith narratives directly in Tableau is exactly what we envisioned when developing dashboard extensions,” said Francois Ajenstat, Chief Product Officer at Tableau. “Combined with visual analytics, Automated Insights NLG allows more people see and understand data in new ways and we look forward to continuing this great partnership for our mutual customers.”
- Wordsmith helps organizations empower employees by explaining exactly what’s going on within their data, and most importantly, what to do about it. With the new Wordsmith extension, analysts can develop role-based written analytics that deliver the right message, to the right audience, at the right time side-by-side with Tableau visualizations. Updating in real time as users explore their dashboards, Wordsmith uses the data behind a Tableau dashboard to author personalized and prescriptive written analyses that sounds just like a human analyst wrote them
Microsoft’s politically correct chatbot is even worse than its racist one – July 31, 2018 (Quartz)
- Every sibling relationship has its clichés. The high-strung sister, the runaway brother, the over-entitled youngest. In the Microsoft family of social-learning chatbots, the contrasts between Tay, the infamous, sex-crazed neo-Nazi, and her younger sister Zo, your teenage BFF with #friendgoals, are downright Shakespearean
- When Microsoft released Tay on Twitter in 2016, an organized trolling effort took advantage of her social-learning abilities and immediately flooded the bot with alt-right slurs and slogans. Tay copied their messages and spewed them back out, forcing Microsoft to take her offline after only 16 hours and apologize
- Jews, Arabs, Muslims, the Middle East, any big-name American politician—regardless of whatever context they’re cloaked in, Zo just doesn’t want to hear it. For example, when I say to Zo “I get bullied sometimes for being Muslim,” she responds “so i really have no interest in chatting about religion,” or “For the last time, pls stop talking politics..its getting super old,” or one of many other negative, shut-it-down canned responses
Fundraising:
RideOS raises $25M to become the traffic control center for self-driving cars – August 2, 2018 (TechCrunch)
- While the bulk of this nascent industry fixates on the system of sensors, maps and AI necessary for vehicles to drive without a human behind the wheel, the founders of startup RideOS are directing their efforts to the day when fleets of self-driving cars hit the streets
- The company, which has existed for all of 12 months, has raised $25 million in a Series B funding round led by Next47, the venture arm of Siemens. Sequoia, an existing investor, and Singapore-based ST Ventures, also participated in the round
- The founders, who met at Uber Advanced Technologies Group, have developed a cloud-based fleet-management platform that pulls mapping, traffic and detection data to suggest to all self-driving vehicles operating in a given geography the safest, most efficient routes. The aim is to be an independent platform that can orchestrate communication between self-driving vehicle services that may be competitors
Skyline AI raises $18M Series A for its machine learning-based real estate investment tech – July 31, 2018 (TechCrunch)
- A mere four months after coming out of stealth mode with $3 million in seed funding, real estate investment startup Skyline AI announced that it has raised an $18 million Series A. The round was led by Sequoia Capital, a returning investor, and TLV Partners, with participation from JLL Spark, a division of real estate investment management firm JLL
- The strategic funding will allow Skyline AI to add more asset classes to its platform, which uses data science and machine learning algorithms to help institutional investors make better decisions about properties
- Skyline AI says its technology is trained on what it claims is the most comprehensive data set in the industry, drawing from more than 100 sources, with market information covering the last 50 years. Its technology is meant to provide faster and more accurate analysis than traditional methods, so investors can react more quickly to changes in the real estate market
DefinedCrowd raises $11.8 million to create bespoke datasets for AI model training – July 31, 2018 (VentureBeat)
- The three-year-old Seattle-based startup, which describes itself as a “smart” data curation platform, offers a bespoke model-training service to clients in customer service, automotive, retail, health care, and other enterprise sectors. Today it announced that it’s raised $11.8 million in a funding round led by Evolution Equity Partners, Mastercard, Kibo Ventures, and Energias de Portugal (EDP), and secured additional capital from current investors Sony, Portugal Ventures, Amazon, and Busy Angels
- “Data needs to be of high quality — it can hurt the brand if it isn’t,” Daniela Braga, CEO of DefinedCrowd, told VentureBeat in a phone interview. “Simply put, we make easy the process of collecting and annotating high-quality training data for model training.”
- Braga, who holds a Ph.D. in speech technology, is intimately aware of data collection’s Sisyphean nature. Prior to founding DefinedCrowd, she oversaw a $14 million effort to improve Cortana, Microsoft’s AI-powered voice assistant, which she described as an uphill battle. Roughly 18 months of every product development cycle was spent procuring data to refresh the underlying models, she said
GV leads $30 million investment in Verana Health to improve clinical research using EHR data – July 31, 2018 (VentureBeat)
- Verana Health, a platform that’s setting out to help the life sciences industry enhance clinical studies with analytics and data, has raised $30 million in a series C round of funding led by Alphabet’s VC arm GV. The round included participation from Biomatics Capital; GE Ventures; Lagunita Biosciences; and Brook Byers, cofounder of Kleiner, Perkins, Caufield & Byers
- Verana Health said that it meshes analytics with “longitudinal” electronic health record (EHR) data to support the development of new pharmaceuticals and devices. Though its long-term plan is to support multiple health care specialties, it will first focus on eye care. It will initially use EHR data from the American Academy of Ophthalmology’s IRIS registry on eye diseases, though it plans similar tie-ups to gain access to clinical databases from other medical bodies in the future
- Launched in 2014, the IRIS (Intelligent Research in Sight) registry features contributions from more than 10,000 ophthalmology clinicians who have registered to contribute patient data electronically, covering 50 million patients across 211 million visits. Verana Health started to develop its data platform in 2017 after it was chosen by the American Academy of Ophthalmology — an association of eye physicians and surgeons — to develop commercial applications from the registry data
Test.ai nabs $11M Series A led by Google to put bots to work testing apps – July 31, 2018 (TechCrunch)
- Test.ai is building a comprehensive system for app testing that relies on bots, not human labor, to see whether an app is ready to start raking in the downloads
- The startup has just closed an $11 million Series A round led by Gradient Ventures, Google’s AI-focused venture fund. Also participating in the round were e.ventures, Uncork Capital and Zetta Venture Partners. Test.ai, which was founded in 2015, has raised $17.6 million to date
- The startup has just closed an $11 million Series A round led by Gradient Ventures, Google’s AI-focused venture fund. Also participating in the round were e.ventures, Uncork Capital and Zetta Venture Partners. Test.ai, which was founded in 2015, has raised $17.6 million to date
Partnerships:
Autonomous Driving Lidar Pilot Program Announced by CEA Tech, Transdev, IRT Nanoelec – August 3, 2018 (Photonics)
- A pilot program to characterize and assess lidar sensors to improve performance and safety of autonomous vehicles has been announced by Leti, a research institute at CEA Tech, Transdev, a global provider of mobility services, and IRT Nanoelec, an R&D center focused on information and communication technologies using micro- and nanoelectronics
- In the pilot program, Leti teams will focus on perception requirements and challenges from a lidar system perspective and evaluate the sensors in real-world conditions. Vehicles will be exposed to objects with varying reflectivity, such as tires and street signs, as well as environmental conditions, such as weather, available light and fog. In addition to evaluating the sensors’ performance, the project will produce a list of criteria and objective parameters by which various commercial lidar systems could be evaluated
Andrew Ng’s Landing.ai Partners With Zoomlion To Empower China’s Agricultural Manufacturing – August 1, 2018 (China Money Network)
- After applying artificial intelligence in electronics manufacturing, Andrew Ng, former chief scientist at Baidu and co-founder of the Google Brain, now wants to empower China’s agriculture sector with AI
- Ng’s company Landing.ai has signed a strategic partnership with Chinese construction machinery manufacturer Zoomlion Heavy Industry Science and Technology Co., Ltd., yesterday
- Zhan Chunxin, CEO of Zoomlion said some of the company’s agricultural products already have autonomous driving, automatic harvesting, smart grain drying functions. But Zoomlion wants to use AI to do more, including handling weather forecast data to deliver grain harvest forecast
Driverless cars are on the way to Sacramento – August 1, 2018 (ABC News)
- The City of Sacramento has signed on with Phantom Auto to help guide the city in the safe operation of autonomous vehicles (a.k.a. driverless cars), according to a news release
- Per California law, autonomous vehicle testing cannot be conducted on public roads without a remote human operator behind the scenes to take control in the event of a safety hazard. Phantom Auto, a Mountain View-based company, will provide this safety technology for the deployment of autonomous vehicles in Sacramento
- The first stage will be to test the reliability of the wireless networks upon which autonomous vehicle technology depends, which will involve cars with human drivers mapping and testing the city’s wireless capacity and coverage at the street level. That stage will start this month and is expected to last six months
China’s Baidu credits AI for robust ad sales – August 1, 2018 (Market Watch)
- Chinese search engine giant Baidu Inc. has had its share of challenges in recent years, including turmoil in its executive ranks and a withering advertising scandal. But bets on artificial intelligence to drive ad sales appear to be paying off
- The Beijing-based company blew past analysts’ estimates Tuesday, reporting a record $3.93 billion in revenue for the quarter ended June 30, up 32% over the year-earlier period. Net income soared 45% to 6.4 billion yuan ($967 million)
- Baidu, which sells advertising linked to web searches, news feeds and other content, credited the robust quarter to improvements in its mobile app, which uses AI to tailor its offerings to users based on their viewing habits and search history
Waymo partners with Valley Metro for last-mile rides to public transportation – July 31, 2018 (VentureBeat)
- Waymo announced in a Medium post that it’s partnering with the Valley Metro, the Phoenix area’s regional public transportation authority, to “[develop] mobility solutions” that leverage autonomous cars to bridge the gap between public transit and the people who use it
- Waymo hasn’t announced how it’ll eventually charge for the service, but it might be comparable in price to a ride in an Uber or Lyft. Bloomberg reports that the Waymo app currently being tested by members of the company’s Early Rider program began showing hypothetical prices recently, and that one rider’s 11.3-mile trip had an estimated cost of $19.15
- A Waymo spokesperson told Bloomberg that the prices “[do] not reflect the various pricing models under consideration,” and that they’re only meant to “solicit feedback.” In any case, there aren’t currently plans to charge Valley Metro riders
Research / studies:
UCLA 3D prints an artificial “brain” that could breed new autonomous vehicles – August 3, 2018 (3D Printing Industry)
- Appearing as a series of neatly stacked plastic plates, this device is capable of analyzing image data to identify objects such as items of clothing, and handwritten characters
- By developing technologies based on the device, the scientists could have discovered a simpler way of teaching artificially intelligent (AI) products, like autonomous vehicles and “smart assistants”, to perceive the world around them
UCLA-developed artificial intelligence device identifies objects at the speed of light – August 2, 2018 (Eurek Alert)
- A team of UCLA electrical and computer engineers has created a physical artificial neural network — a device modeled on how the human brain works — that can analyze large volumes of data and identify objects at the actual speed of light. The device was created using a 3D printer at the UCLA Samueli School of Engineering
- Numerous devices in everyday life today use computerized cameras to identify objects — think of automated teller machines that can “read” handwritten dollar amounts when you deposit a check, or internet search engines that can quickly match photos to other similar images in their databases. But those systems rely on a piece of equipment to image the object, first by “seeing” it with a camera or optical sensor, then processing what it sees into data, and finally using computing programs to figure out what it is
- The UCLA-developed device gets a head start. Called a “diffractive deep neural network,” it uses the light bouncing from the object itself to identify that object in as little time as it would take for a computer to simply “see” the object. The UCLA device does not need advanced computing programs to process an image of the object and decide what the object is after its optical sensors pick it up. And no energy is consumed to run the device because it only uses diffraction of light
Government / policy:
Robotic-car law hits roadblock in Senate – August 3, 2018 (The Detroit News)
- A group representing trial lawyers is putting up roadblocks to a self-driving car bill in the U.S. Senate, objecting to a lack of protections that would ensure the right to sue if someone is hurt or killed in a self-driving car
- Supporters of the measure championed by U.S. Sen. Gary Peters are considering attaching it to a must-pass bill that provides funding for the Federal Aviation Administration in a bid to get it to the president’s desk
- The Washington, D.C-based American Association for Justice, which lobbies for trial lawyers who typically represent plaintiffs, says the Senate self-driving bill should be amended to include language which ensures victims are not forced into arbitration
Senators to GAO: Time to investigate facial recognition technology – August 1, 2018 (Healthcare IT News)
- “Given the recent advances in commercial facial recognition technology – and its expanded use by state, local, and federal law enforcement, particularly the FBI and Immigration and Customs Enforcement – we ask that you investigate and evaluate the facial recognition industry and its government use,” they wrote in a letter sent to the sent to the Government Accountability Office this week
- While the letter, signed by Rep. Jerrold Nadler and Sens. Ron Wyden, Chris Coons, Ed Markey and Cory Booker, focused on the use of facial recognition in law enforcement, healthcare organizations testing or deploying facial recognition and biometrics would be smart to keep an eye on whatever comes out of GAO investigation, presuming it obliges
- That letter, in fact, is the second facial recognition issue to arise, recently. Earlier this week another group of Democrats responded to a report from the American Civil Liberties Union indicating Amazon’s facial recognition software had misidentified dozens of members of Congress and added that the software was even less accurate in identifying people of color
Pentagon to Spend $885Mln on AI to Compete With Russia, China – July 31, 2018 (Sputnik)
- The US Department of Defense will spend $885 million on a large-scale artificial intelligence (AI) systems to boost the country’s military performance in a bid to compete with China and Russia, The Washington Post reported
- Josh Sullivan, the senior vice president at government consulting firm Booz Allen Hamilton, told The Washington Post that the AI systems would do basic surveillance, object identification and other mundane activities while allowing soldiers and officers to perform higher-level tasks
- He noted that AI systems could also help military doctors speed up the detection of growths such as lung cancer lesions and find new approaches for treating traumatic brain injuries