What’s happened in AI: June 18th-24th

This week we saw another tech giant face internal push back on partnering with the government on AI initiatives. It was Google over the past few weeks with Project Maven, and now it’s Amazon with the use of its facial recognition software with U.S. police departments. Will be interesting to see if this trend continues.

On the fundraising side, AI chip startup Cambricon led the way with a Series B that valued the company at ~2.5bn. Their investors are some of China’s most influential funds, tech startups, and tech giants. Cambricon is definitely a company to keep on your watch list.

Company developments:

Woman in self-driving Uber was watching The Voice before fatal crash in Phoenix, Arizona – June 23, 2018 (ABC News)

• Arizona police say the woman riding inside a self-driving Uber that crashed and killed a pedestrian in Arizona earlier this year was watching The Voice on her phone at the time
• That might seem like a reasonable thing to do inside an autonomous car, but police say the crash could have been avoided if she had not been distracted, and are looking at a vehicular manslaughter charge
• Video taken from inside the vehicle shows Ms Vasquez looked down 204 times during the 1.6-kilometre journey, and only looked up about half-a-second before the Volvo, travelling at 70 kilometres an hour, hit Ms Herzberg

Samsung Galaxy S10 to come with facial recognition and in-display fingerprint scanner – June 23, 2018 (Deccan Chronicle)

• Over the past few weeks, there have been a lot of leaks and rumours surrounding Samsung’s upcoming flagship handset the Galaxy Note 9. The last we heard about it was that there was a possibility that it would come loaded with a triple-camera system like the Huawei P20 Pro
• This time around, the rumour mill has churned out some news about Samsung’s S-lineup of handsets. As per a report by GSMArena, it appears that Samsung is going to ditch the iris scanner and the Galaxy Note 9 will be the last handset to support this technology. The report notes that in place of the iris scanner, Samsung will invest in a new 3D facial recognition technology. It is believed that Samsung is in the works with Israeli-based firm, Mantis Vision to develop the tech needed to implement the facial recognition

China’s Baidu tests driverless cars on expressway – June 22, 2018 (Economic Times)

• Chinese artificial intelligence (AI) giant Baidu Inc. has tested its two self-driving cars on an unused expressway in Tianjin City, the company said. It was the first time for a Baidu driverless car to be tested, the company said after the exercise on Thursday
  The trial run would enable developers to collect data on the cars’ performance in positioning, sensing the environment and self control, a senior engineer with the road test team said

Microsoft matches Google Lens with AI-powered visual search for Bing – June 22, 2018 (The Verge)

• Microsoft has launched a new “visual search” function for Bing, which lets users snap a picture of something with their phone to search for it online. The feature looks very similar to Google Lens and offerings we’ve seen from third parties that leverage the power of AI to perform quick and accurate object recognition on photos
• The new visual search is available in a range of apps, including the standalone Bing app on iOS and Android, Microsoft Launcher on Android, Microsoft Edge on Android, and coming soon to Edge on iOS. You can use it to search for things like breeds of dogs, famous landmarks, and even items of clothing. The app will try to identify what it sees and, in the case of consumer goods, find places to buy them online

Adobe is using machine learning to make it easier to spot Photoshopped images – June 22, 2018 (The Verge)

• The company’s latest work, showcased this month at the CVPR computer vision conference, demonstrates how digital forensics done by humans can be automated by machines in much less time. The research paper does not represent a breakthrough in the field, and it’s not yet available as a commercial product, but it’s interesting to see Adobe — a name synonymous with image editing — take an interest in this line of work
• Speaking to The Verge, a spokesperson for the company said that this was an “early-stage research project,” but in the future, the company wants to play a role in “developing technology that helps monitor and verify authenticity of digital media.” Exactly what this might mean isn’t clear, since Adobe has never before released software designed to spot fake images. But, the company points to its work with law enforcement (using digital forensics to help find missing children, for example) as evidence of its responsible attitude toward its technology
• The new research paper shows how machine learning can be used to identify three common types of image manipulation: splicing, where two parts of different images are combined; cloning, where objects within an image are copy and pasted; and removal, when an object is edited out altogether

SAS named as a Leader in AI-based text analytics by independent research firm – June 22, 2018 (Business Insider)

• SAS® Visual Text Analytics uses intelligent algorithms and natural language processing (NLP) techniques to automatically extract relationships and patterns within unstructured data, thereby eliminating the need for manual analysis. The NLP tools help users in sentiment analysis, speech to text, natural language understanding and natural language generation
• SAS earned top marks in the Forrester report in many sub-categories for current offering including: analytics, ease of use, mix of rules and machine learning, depth of functionality, and others
• The report states: “SAS’s brand speaks for itself as a leader in advanced analytics; as a result, SAS Visual Text Analytics comes with a number of machine learning models. Users can also leverage other capabilities of the platform, such as forecasting and optimization, to deliver predictive, prescriptive, and actionable analytics.”

Visa Makes Travel Banking And Card Canceling Easier With Chatbots – June 22, 2018 (PSFK)

• Visa Canada is aiming to make banking easier from a remote setting with artificial intelligence and chatbots. Through a partnership with Finn AI, bank members will be able to contact support services 24/7, flag suspicious transactions, locate nearby ATMs, and even cancel misplaced or lost cards using chat. This function also makes travel easier for Visa users as they can manage travel notifications before and after takeoff and check conversion rates
• The conversational AI is intended to make accessing Visa’s online banking services a more intuitive and pleasant experience. Advanced natural language processing is always accessible, and makes customers feel more welcomed and encouraged to reach out with questions and problems. Visa Canada’s goal is to improve customer support by making interactions with financial institutions as quick and easy as having a conversation with a friend

Narrative Science Awarded as Technology Pioneer by World Economic Forum – June 21, 2018 (Kaplan Herald)

• World Economic Forum on Thursday announced its selection of the world’s 61 most promising Technology Pioneers 2018, the most diverse class of such pioneers ever
• Companies were selected for their potential to “transform society and industry” and “shape the Fourth Industrial Revolution,” Fulvia Montresor of the World Economic Forum said
• Narrative Science, which creates technology that transforms data into human-like language, was selected for its contributions in the field of artificial intelligence

JP Morgan is unleashing artificial intelligence on a business that moves $5 trillion for corporations every day – June 20, 2018 (CNBC)

• For the past six months, J.P. Morgan Chase has quietly been shuttling clients to the bank’s technology lab on the West Side of Manhattan for a glimpse at the future of work
• The executives — treasurers and finance chiefs of some of the world’s biggest companies — were introduced to an artificial intelligence-powered servant that J.P. Morgan hopes will soon be able to answer queries and anticipate their needs, a first in the world of corporate payments. But before that could happen, the machine needed contact with people to figure out what makes humans tick. The bank is widening the trials with select clients next week and plans to release it to more companies next year
• J.P. Morgan’s latest attempt at furthering automation in finance is happening in its treasury services division, a key business that helps corporate clients from Honeywell International to Facebook move money around the world. Unseen by retail consumers, the unit handles an average of $5 trillion daily, from the mundane (payroll and supplier remittances) to the unusual (multibillion-dollar checks for huge mergers)

Singapore firm to launch ‘AI doctor’ in China – June 20, 2018 (SGSME)

• An artificial intelligence (AI) machine developed by a Singapore firm called Biomind is making the tedious process of diagnosing a brain tumour easier and less prone to human error. The machine can do in a second what an experienced doctor would take 30 minutes to complete, given the thousands of magnetic resonance imaging (MRI) images that need to be checked
• It was developed at a research centre jointly run by Hanalytics and Beijing Tiantan Hospital, where it has been used on patients over the past two months. China is trying to become a world leader in AI and has been moving aggressively to adopt it in the healthcare sector
• Biomind can pick up neurological conditions with an accuracy of over 90 per cent, higher than the 60 per cent by top doctors at the Beijing hospital, noted Mr Moh. The machine is linked to the hospital’s MRI and CT scanners, so doctors can opt to have it process brain scans

Google injects Hire with AI to speed up common tasks – June 19, 2018 (TechCrunch)

• Since Google Hire launched last year it has been trying to make it easier for hiring managers to manage the data and tasks associated with the hiring process, while maybe tweaking LinkedIn while they’re at it. Today the company announced some AI-infused enhancements that they say will help save time and energy spent on manual processes
• “By incorporating Google AI, Hire now reduces repetitive, time-consuming tasks, like scheduling interviews into one-click interactions. This means hiring teams can spend less time with logistics and more time connecting with people,” Google’s Berit Hoffmann, Hire product manager wrote in a blog post announcing the new features
• Streamlined processes include scheduling interviews, reviewing resumes, and making phone calls to potential candidates

Engineer.ai aims at build-an-app platforms by using AI for speed and cost – June 19, 2018 (TechCrunch)

• Engineer.ai has been up and running in private beta mode for two and a half years, totally boo-strapped by its founders. But today it launches with the claim to be the world’s first human-assisted AI for building custom digital products. It’s Builder platform claims to combine artificial intelligence with crowdsourced teams of designers and developers to build bespoke digital products at – they say – twice the speed and less than a third of the cost of traditional software development
• So were are talking the ability to build a smartphone app, say, very easily and quickly, choosing from recommended features or adding others. The AI then creates a “build card” which guarantees a maximum price and estimated delivery date. It then taps a library of existing components and manages crowdsourced global teams for aspects that might be unique. The platform can host the end product and carry on upgrading the product
• Co-founded by serial entrepreneurs Sachin Dev Duggal and Saurabh Dhoot, the company, which has been bootstrapped on the proceeds of the sale of their previous startup, nivio. Virgin Unite, Richard Branson’s philanthropy arm, used it for its Campaign against the Death Penalty & Criminal Justice, for instance

Yelp uses machine learning to create Popular Dishes list – June 19, 2018 (The Next Web)

• Yelp’s latest update will compile a list of Popular Dishes for each restaurant, using AI to figure out what everyone prefers to eat based on the restaurant’s reviews
• Users can see a list of Popular Dishes on each restaurant’s profile. Tapping on the list will bring you to a more detailed page with photos, price, and menu listing. For each dish, it shows relevant reviews mentioning the food’s quality, for good and for ill. According to a Yelp spokesperson, the reviews don’t currently take positive or negative associations with the food into account — it’s purely based on how many times they’re mentioned in the reviews

Sherpa, the Spanish personal assistant, launches an API for its predictive AI with Porsche as 1st customer – June 19, 2018 (TechCrunch)

• Sherpa, a personal assistant startup that has carved out a niche for itself by focusing on the Spanish-language market (alongside English) and predictive suggestions, is expanding. The company is launching a set of APIs called the Sherpa Platform, which will let other businesses tap into its predictive recommendations and use them in their own consumer-facing services.
• Sherpa — based out of Bilbao, Spain and Palo Alto — is also announcing its a customer for the service: Porsche, which plans to use the service in its connected car services in its luxury vehicles. In addition to the automotive sector, Sherpa plans to target the home and mobile segments with its Platform APIs, and it has some deals specifically with other automotive companies and telecoms carriers in the works
• Sherpa is also in the process of raising money. So far its raised a modest $6.5 million of funding, from Alma Mundi Innvierte Fund, FCRE, and unnamed private investors (“celebrities” says Uribe-Extebarria). The company is close to completing a bridge round for later this year of around $8 million, ahead of a larger Series B. Uribe-Extebarria says he has spoken to “all the usual names” in the US — he splits his time between Spain and California — and also a number of investors in Europe

Ford’s plan to turn Detroit’s oldest neighborhood into an electric, AV hub – June 19, 2018 (TechCrunch)

• Ford will spend the next four years transforming at least 1.2 million square feet of space in Corktown — Detroit’s oldest neighborhood — into a hub for its electric and autonomous vehicles businesses
• The centerpiece of the new Corktown campus will be the long-abandoned Michigan Central Station, which Ford acquired

Amazon Shareholders Ask Company Not to Sell Facial Recognition Software to Government – June 19, 2018 (Government Technology)

• A coalition of nearly 20 groups of Amazon shareholders have joined together to send a message to CEO Jeff Bezos: Stop selling Rekognition software to police
• In the letter, the shareholders express their concerns about how the software can be used. Right now, users can use the software to detect objects and faces in images and videos and upload a face to automatically identify individuals. In the letter, the shareholders wrote that, “We are concerned the technology would be used to unfairly and disproportionately target and surveil people of color, immigrants, and civil society organizations,” and “We are concerned sales may be expanded to foreign governments, including authoritarian regimes.”
• There are those at Amazon who think that putting a ban on the sales of the software would be pointless since Amazon’s policy prohibits the use of its software for activities that are illegal, violate the rights of others, or may be harmful in any way. “We believe it is the wrong approach to impose a ban on promising new technologies because they might be used by bad actors for nefarious purposes in the future. The world would be a very different place if we had restricted people from buying computers because it was possible to use that computer to do harm,” said Matt Wood, general manager of artificial intelligence at Amazon Web Services in a blog post

Google’s health AI model can predict when a patient will die – June 19, 2018 (Digital Journal)

• In trials that used data from two U.S. hospitals, the researchers were able to show that the algorithms in the model improved predictions as to the length of stay and time of discharge — but also the time of death of patients. The neural network in the study uses an immense amount of data, including information about the patient’s medical history and vitals. A new algorithm lines up events in the patient’s records into a timeline, and this enables the learning model to predict future outcomes such as length of stay and even death. The predictions are made in record time as well
• The predictions may enable the hospital to prioritize patient care, or adjust treatment plans, and even recognize medical emergencies before they happen. It will also free up the time of healthcare workers who had to sift through all this data in order to make less accurate predictions
• Some worry that collecting this massive amount of personal health data and putting it all into a single model that is owned by Google, one of the largest private companies in the world, could very well create problems. There would need to be guarantees that the data would not be shared. Many companies would be anxious to use the information for their purposes. As the article notes: ” Electronic health records of millions of patients in the hands of a small number of private companies could quickly allow the likes of Google to exploit health industries, and become a monopoly in healthcare.”

Chinese search giant Baidu creates an open-source A.I. for detecting cancer – June 18, 2018 (Digital Trends)

• The company has developed an A.I. that is capable of analyzing slides containing biopsied tissue. Reviewing these slides can be difficult, even for experienced pathologists, but Baidu’s deep learning technology is able to look for tiny tumor cells faster and with greater accuracy than previous approaches. In tests, the algorithm was able to outperform both a professional pathologist and the previous winner of the so-called Camelyon16 challenge, a competition intended to evaluate algorithms for automated detection of cancer metastasis in lymph node tissue sections
• “Using A.I. to analyze pathology images is a very challenging task,” Yi Li’s, a machine learning research scientist at Baidu, told Digital Trends. “A digitized pathology slide at 40x magnification often contains billions of pixels, which is too large for a neural network to process. As a result, the mega-image is divided into tens of thousands of smaller individual images so that a neural network can analyze each of them separately. What’s unique about our neural conditional random field (NCRF) algorithm is that it can look at multiple images — including the potentially cancerous region and its surroundings, simultaneously. This new capability significantly reduces the number of false positives [in the form of] misclassified normal cells.”

IBM shows off an artificial intelligence that can debate a human – and do pretty well – June 18, 2018 (USA Today)

• At a media gathering here Monday afternoon, a black, artificial intelligence-infused IBM computer with a screen for a face more than held its own debating seasoned human debaters
• In one debate faceoff, IBM’s “Project Debater” AI computer made the case in favor of the government subsidizing space exploration against Israeli debate champion Noa Ovadia, who took the opposite position. Ovadia was judged the winner by the crowd of journalists in “delivering” the argument – the computer’s attempts at humor didn’t measure up to the personality of a human – but IBM outscored Ovadia handily on the question of “knowledge enrichment
• IBM’s computer fared better in a second debate persuading the crowd that telemedicine is worth pursuing against another human debater, Dan Zafrir. Again, the human prevailed on delivery, but this time only by a slim margin, and the computer was a big winner in knowledge enrichment. And at least nine audience members’ minds on the topic changed to the point of view of the computer

M&A:

PayPal to buy Simility, a specialist in AI-based fraud and risk management, for $120M – June 21, 2018 (TechCrunch)

• Payment provider PayPal continues apace with its acquisitions streak to bring more modern tools into its platform to serve its 237 million customers. Today the company announced that it is buying Simility, a fraud prevention specialist, for $120 million in cash
• PayPal had been an investor in Simility (it owns three percent of the company, it says), along with Accel, the Valley Fund, Trinity Ventures and others. The startup had raised just under $25 million and was last valued at $52.75 million, according to figures from PitchBook, making this a decent return for its backers. The deal is expected to close in Q3
• Simility’s approach is to use a set of APIs and beacons that essentially monitor digital transactions and buying activity wherever they happen to take place: on mobile, web or in physical environments. Augmenting these with machine learning and feeds from other data sources, it creates something it calls “adaptive” risk management: a changing approach and protection strategy based on what the threat of the moment might be

Microsoft is buying AI startup, Bonsai – June 20, 2018 (TechCrunch)

• If all of the big tech co’s agree on one thing at the moment, it’s that artificial intelligence and machine learning point the way forward for their businesses. As a matter of fact, Microsoft is about to acquire Bonsai, a small Berkeley-based startup it hopes to make the centerpiece of its AI efforts
• The company specializes in reinforcement learning, a kind of trial and error approach to teach a system within the confines of a simulation. That learning can be used to train autonomous systems to complete specific tasks. Microsoft says the acquisition will serve to forward the kind of research the company has been pursuing in the field by leveraging its Azure cloud platform
• “To realize this vision of making AI more accessible and valuable for all, we have to remove the barriers to development, empowering every developer, regardless of machine learning expertise, to be an AI developer,” Microsoft Corporate VP Gurdeep Pall said in an announcement. “Bonsai has made tremendous progress here and Microsoft remains committed to furthering this work.”

WalkMe buys Israeli machine learning co DeepUI – June 20, 2018 (Globes)

• Israeli website navigation and digital adaptation platform company WalkMe today announced that it has acquired Israeli startup DeepUI, a company in stealth mode that has developed a patented machine learning technology to understand any business software at the graphical user interface (GUI) level, without the need for an application programming interface (API). No financial details were disclosed
• The company has 600 employees with over 300 in Israel. DeepUI is WalkMe’s third acquisition since the company was founded in 2011. WalkMe acquired native mobile AI startup Abbi in January 2017, and visual analytics startup Jaco in April 2017. Both technologies have been fully integrated into WalkMe’s Digital Adoption Platform, delivering a comprehensive solution to understand user behavior and simplify the user experience
• Founded in 2014, DeepUI’s co-founders, Dr. Ron Zohar and Moran Shemer, have over 35 years’ combined experience in the design and development of advanced algorithmic systems. The DeepUI team will join WalkMe to help automate virtually every facet of the digital user experience, ensuring technology return on investment while continuing to improve employee productivity

Fundraising / investment:

Agari raises $40 million to grow its AI-powered email authentication platform – June 22, 2018 (VentureBeat)

• It has been a big week for cybersecurity investments, and the latest such startup to announce a notable tranche of funding is California-based Agari, which uses predictive artificial intelligence (AI) to thwart email attacks. Founded in 2009, Agari’s platform detects potential email breaches, identifies malicious inbound messages, and establishes the authenticity of a sender before delivering an email to its intended recipient. The company claims a number of big-name clients, including Facebook, Google, and Microsoft
• Agari has announced a fresh $40 million in financing, led by Goldman Sachs Growth Equity. Participation from existing investors includes Norwest Venture Partners, Scale Venture Partners, Battery Ventures, Greylock Partners, First Round Capital, and Alloy Ventures. Prior to now, the company had raised $48 million, and with another chunk of cash in the bank Agari plans to expedite its growth in Europe and Asia
• “Safeguarding against sophisticated email deception attacks and account takeover has become a boardroom-level mandate, with business email compromise costing companies billions of dollars each year,” noted Agari CEO Ravi Khatod. “This investment led by Goldman Sachs will fund a new, critical growth phase for Agari, as we deepen our product, data science, and go-to-market investments to deliver success to every Agari customer and partner globally.”

AI chip manufacturer Cambricon closes Series B at $2.5bn valuation – June 21, 2018 (Global Manufacturing)

• Chinese chip manufacturer, Cambricon Technologies, has secured hundreds of millions in its latest funding round, reports Synced Review. The Series B funding round has left the company with a valuation of $2.5bn. Investors reportedly include Capital Venture (a Chinese state-owned company), SDIC Venture Capital, China Capital Investment Group, TCL Capital, CITIC Securities, Alibaba VC, and Lenovo Capital & Incubator Group. China Daily says that after the funding round, it is now the highest valued smart chip startup in the world.
• Cambricon mainly focusses on the development of artificial intelligence (AI) chips. The company produced China’s first AI chip in 2016 for use in smartphones, wearable technology and drones. In particular, Cambricon is looking to develop chips with low power consumption
• Previous investors in Cambricon have included iFlyTek, Oriza and Yonghua, according to Global University Venturing. iFlyTek is a voice recognition company and part of China’s national AI team, while both Oriza and Yonghu are early-stage investment companies with a technology focus

Silexica, which optimises how disparate applications work together on autonomous cars, raises $18M – June 21, 2018 (TechCrunch)

• Silexica, a startup based out of Cologne, Germany, that has developed a set of tools to help map and optimise a wide range of applications across multicore processors — specifically the kinds of applications and computers that power self-driving cars — has raised $18 million
• The Series B is led by EQT Ventures, the newish firm that sits under the PE firm EQT Partners, based out of Stockholm. Existing investors Merus Capital, Paua Ventures, Seed Fonds Aachen (Silexica originally spun out of the University of Aachen) and DSA Invest all also participated in this round. The company has raised $28 million to date
• The funding, Silexica’s CEO Maximilian Odendahl (who co-founded the team with Johannes Emigholz and Weihua Sheng) told TechCrunch, will be used not just to increase its optimisation features — allowing for an increasing variety of services and functionalities to be monitored and diagnosed — but also to add a cloud component to how it monitors and processes information on a vehicle. Customers include Denso, Toyota, Fujitsu and Huawei — underscoring the range of potential buyers of its tech

Vi raises $20 million to expand its AI fitness business – June 19, 2018 (VentureBeat)

• New York-based startup Vi today announced that it has raised $20 million to expand its artificially intelligent fitness business. Investors include Joy Ventures, Square Peg Capital, New Era Ventures, Cerca Partners, Triventures, Wellborn Ventures, and FGI Capital Partners
• “With this new round of funding, Vi will be focusing on two main areas,” Vi CEO Omri Yoffe told VentureBeat via email. “First, we will be doubling down on Vi’s personalization, real-time feedback, and awesome, intelligent voice/audio content. Second is to expand Vi’s offering in terms of additional fitness and wellness activities. Our first big announcement following this round is on near horizon with new features [and] will take place mid-July.”
• Vi sells the eponymous $149 Vi earbuds, which raised more than $1.68 million on Kickstarter in 2016. They feature Harman Kardon sound, integration with Spotify and Apple Music that queues up tracks matching the beats per minute of the recommended pace, and an AI-powered running coach that analyzes performance data and provides feedback in real time. (If you’re jogging slower than usual, for example, Vi will politely nudge you to pick up the pace.)

Photomyne raises $5 million for its A.I.-powered photo scanning app – June 18, 2018 (TechCrunch)

• Tel Aviv-based Photomyne, an A.I.-powered app that helps you bring your old photo prints online, has been benefiting from the subscription app boom to the tune of $5 million in Series A funding. Today, the app is used by a million people every month, and 250,000 people pay the $20 annual subscription for the expanded service. This adds a handful of additional features, including the option to build a family website where all your photos are uploaded immediately after being scanned
• There is something of a limited lifetime for apps that convert physical media to digital – at some point, everyone who wants to transition their old media to the web will have done so. Another issue is that some people will make scanning photos a one-time project. They’ll then save all their photos to their own device and cloud storage, and cancel their subscription
• The Series A round was led by Luxembourg-based Maor, a co-investment tech fund from Philippe Guez and Eric Elalouf. It also included participation from Israeli investors and others from its seed round a couple of years ago

Partnerships:

Waymo, Uber, Ford, and others are joining forces to explore the ‘human impact’ of self-driving cars – June 19, 2018 (The Verge)

• The Partnership for Transportation Innovation and Opportunity (PTIO) is a newly formed group comprised of most of the major companies that are building and testing on self-driving cars. This includes legacy automakers like Ford, Toyota, and Daimler; tech giants like Waymo (née Google), Uber, and Lyft; and logistics providers like FedEx and the American Trucking Association. The new organization is being formed as a 501(c)(6), which allows it to accept donations like a nonprofit and lobby government like a chamber of commerce
• “Concern for the safety of workers and the public is paramount to PTIO,” the group’s executive director Maureen Westphal, said in an email to The Verge, “and safe deployment of [autonomous vehicle] technology is fundamental to securing better job opportunities for workers, so we plan to engage with a variety of concerned stakeholders already having conversations and planning for this transition to an autonomous vehicle future.”
• The group has its work cut out for it. According to the Bureau of Labor Statistics, over 3.8 million people operate motor vehicles for their livelihood. This includes truck driving, the most common profession in 29 US states, which employs about 1.7 million people. Many expect self-driving trucks to be among the first autonomous vehicles to hit the road. When autonomous vehicle saturation peaks, US drivers could see job losses at a rate of 25,000 a month, or 300,000 a year, according to a report from Goldman Sachs Economics Research released last year

Research / studies:

Salesforce Research aims to capture the nuances of natural language processing – June 22, 2018 (SD Times)

• The Salesforce Research team is attempting to capture the nuances of natural language processing with a new generalized model. The team described its approach in a recently published a paper on the Natural Language Decathlon (decaNLP)
• According to Richard Socher, chief scientist at Salesforce who is leading the research team, while natural language processing is opening up new opportunities for machines, language understanding is becoming difficult because it connects and requires other areas of intelligence such as visual, emotional and logical areas
• The model is designed to tackle 10 different NLP tasks at once and eliminate the need to build and train individual models for each NLP problem

Species-identifying AI gets a boost from images snapped by citizen naturalists – June 21, 2018 (TechCrunch)

• Researchers have assembled hundreds of thousands of images taken by regular folks of critters in real life situations — and by studying these, our AI helpers may be able to get a handle on biodiversity
• Many computer vision algorithms have been trained on one of several large sets of images, which may have everything from people to household objects to fruits and vegetables in them. That’s great for learning a little about a lot of things, but what if you want to go deep on a specific subject or type of image? You need a special set of lots of that kind of image
• The researchers, from Caltech, Google, Cornell and iNaturalist itself, put together a limited subset of the more than 1.6 million images in the app’s databases, presented this week at CVPR in Salt Lake City. They decided that in order for the set to be robust, it should have lots of different angles and situations, so they searched for species that have had at least 20 different people spot them

UC Irvine researchers created an AI that can solve a Rubik’s Cube in 30 moves – June 19, 2018 (Mobile Syrup)

• A research team from the University of California, Irvine (UC Irvine) created an artificial intelligence that taught itself to solve a Rubik’s Cube in 30 moves. The team was comprised of PhD candidates Stephen McAleer, Forest Agostinelli and Alexander Shmakov, as well as UC Irvine computer science professor Pierre Baldi. Their paper can be read online through Cornell University’ arXiv preprint server. The paper outlines the process behind creating and training the DeepCube AI
• To overcome the limitations of reinforcement learning, the team created and utilized an AI technique known as Autodidactic Iteration. This allows an AI to create its own reward system, so that it can then use reinforcement learning
• It only took DeepCube 44 hours to learn how to complete solve any standard, randomly scrambled Rubik’s Cube in an average of 30 moves. That puts DeepCube right amidst the speedcubers who employ human historical knowledge

MIT’s “smart power outlets” use machine learning to prevent circuits tripping unnecessarily – June 18, 2018 (New Atlas)

• Arc-fault detectors can be invaluable protectors against electrical fires, but they can also be a nuisance if they start switching off devices and appliances when there’s no safety risk. To cut down on the false alarms, an MIT team of engineers is working on new “smart outlets” that use neural networks to not only determine if a detected arc is dangerous, but also what type of devices are plugged into the outlet
• They did this using a Raspberry Pi Model 3 microcomputer to monitor the incoming electrical data through an inductive current clamp installed around the outlet wire to sense the current. The clamp runs what it senses through a standard USB sound card with an integrated memory buffer, which is capable of handling very small signals at very high rates of 48 kHz, or 48,000 times a second
• Over time, the network is able to learn as more data is analyzed, so it is increasingly capable of accurately telling which signals to act on and which to ignore, as well as being able to tell which devices are plugged into it. The idea is that one day such smart outlets will operate in many people’s homes to protect electronic devices, while a smartphone apps would allow consumers to analyze and share data about their electrical usage on a very detailed level. By anonymously sharing this data, the outlets will be able to further refine their algorithms

Government / policy:

Roundabouts and potholes could put the brakes on driverless cars in Britain – June 23, 2018 (The Telegraph)

• The crumbling conditions and quirks of the British road system are proving to be a roadblock for autonomous vehicle (AV) companies, according to a senior technician at FiveAI
• “One thing that would definitely speed up autonomous cars getting onto the roads would be having more money spent on fixing potholes,” said John Lusty, FiveAI’s new head of simulations who left his job at Facebook to join the start-up

Orlando International becomes 1st U.S. airport to commit to using facial recognition technology – June 22, 2018 (WPSD Local 6)

• Orlando International has become the first U.S. airport to commit to using facial recognition technology. It involves using facial biometric cameras that can be put near places like departure gates. The cameras verify a passenger’s identity in less than two seconds, and they’re 99 percent accurate
• Facial recognition is also being tested at 13 other U.S. airports

Boston to expand autonomous vehicle testing programme city-wide – June 22, 2018 (Smart Cities World)

• The City of Boston has announced that autonomous vehicle software developer nuTonomy (an Aptiv company), has been authorised to expand testing on Boston streets city-wide
• Boston is using a mandatory and phased approach for its autonomous vehicle testing programme. Under the City’s supervision, nuTonomy has been testing on streets in the Seaport District since January 2017
• Creating a policy for the operation of autonomous vehicles on Boston streets is one of the priorities of the Go Boston 2030 Transportation Plan. The City sees the technology as having great potential to significantly enhance mobility for Boston residents, particularly senior citizens and people with disabilities

TARTA bringing autonomous vehicle program to Toledo next year – June 21, 2018 (WTOL)

• Toledo Area Regional Transit Authority (TARTA) on Thursday announced plans to bring autonomous vehicles to Toledo next year as part of its Move Toledo program. The test program is to start next year and is coming via an $1.8 milion grant from the Federal Highway Administration to test the autonomous buses in the city
• TARTA secured the grant in collaboration with the Ohio Department of Transportation and the Toledo Metropolitan Area Council of Governments. The grant is expected to fully fund the program; no local tax dollars will be used
• The route will take passengers along circuit yet to be chosen in Toledo and will have a driver present for safety issues.
  The test period is for three years

China to build national AI-assisted translation platform – June 20, 2018 (Xinhua.net)

• Chinese AI giant iFlytek Wednesday announced a plan to work with China International Publishing Group (CIPG) to build a national translation platform based on its AI technologies. Specializing in machine translation, iFlytek has been providing real-time Chinese-to-English translation and interpretation services for government organizations and companies both in China and abroad.”The latest machine can translate Chinese to 33 languages. Our technologies in voice recognition, machine translation, and voice synthesis are leading in the world,” said Liu Qingfeng, president of iFlytek
• According to Fang Zhenghui, the deputy director of CIPG, AI translation devices can lift the heavy burden off human translators. Human translators can barely meet the ever-growing translation needs of government departments and Party offices at all levels and Chinese enterprises operating abroad, said Fang
• In the future AI translation platform, CIPG will open its corpus and database to iFlytek so as to improve the accuracy of machine translation

Fastest driverless vehicle to hit Australian roads unveiled in Adelaide – June 19, 2018 (ABC News)

• The fastest driverless vehicle to hit Australia’s roads has been unveiled in Adelaide in a five-year trial to transport university students in the southern suburbs
• The Flinders Express, or FLEX, will initially operate as a shuttle between Flinders University and a nearby train station, before expanding its route within a year to include the nearby hospital
• It is the first time a driverless vehicle has been allowed on public roads in South Australia, with the State Government giving special permission for the university to proceed with the trial

Federal Regulators Tout Autonomous Vehicles’ Potential Benefits – June 19, 2018 (Transport Topics)

• Commercial vehicles utilizing automated technology have the potential to increase highway safety and provide a boost to the economy, the country’s top trucking regulator said June 19. “This may very well be one of the tools that will lead us to safer roadways,” Federal Motor Carrier Safety Administration chief Ray Martinez said during a forum on autonomous vehicles at the University of Michigan’s Mcity, a research and testing site
• The agency is seeking input from stakeholders as it readies an advance notice of proposed rulemaking for commercial vehicles equipped with automated driving systems. Publication of the notice is scheduled for December
• Specifically, FMCSA is aiming to learn from stakeholders whether drivers would be needed inside autonomous trucks operating on the roadways, the challenges of developing the technology around the country and the requisite technology to safely deploy autonomous trucks

Events:

LG CEO , CTO to outline vision for ThinQ Artificial Intelligence at IFA – June 24, 2018 (Vanguard)

• Ahead of IFA world’s leading trade show for consumer electronics and home appliance slated to hold this August in Berlin, Germany, LG Electronics Chief Executive Officer, Jo Seong-Jin and Chief Technology Officer, Dr. I.P. Park will deliver a joint opening keynote at the trade exhibition to outline their vision for LG’s ThinQ strategy for Artificial Intelligence
• With interest in all things AI at a peak, CEO Jo, according to LG Electronics would provide his insight into how LG’s AI strategy will change customers’ lives based on its unique philosophy of an “open platform, open partnership and open connectivity”