There were a few interesting developments this week in various areas of AI. On the autonomous vehicle front, GM’s Cruise is now expanding to Seattle. Ultimately they’re aiming to hire up to 200 engineers for the office there. Seems like GM is really making a big push this year to expand across the U.S. with their autonomous vehicle program.
Meanwhile, one of the most high profile AI events will be moving to Africa in 2020 due to visa issues for many of the attendees. Yoshua Bengio (founder of Element AI) made the call this past Saturday for the International Conference on Learning Representations (ICLR). He noted that many researchers from the Black in AI group where particularly facing visa difficulties. This is a real positive step for fostering an inclusive and diverse environment for AI research. Other weekly news can be found below.
Company developments:
Amazon Transcribe gains live audio transcription feature – Nov. 20, 2018 (VentureBeat)
- The live audio transcription feature is generally available this week and enables developers to pass streams to Transcribe and receive text transcripts in real time. As Paul Zhao, senior product manager at AWS’ machine learning division, and Paul Kohan, senior software engineer at Amazon Transcribe, explained in a blog post, it leverages data-transporting protocol HTTP/2 to transmit audio and transcripts between apps and Transcribe — specifically, HTTP/2’s bidirectional streams implementation, which lets apps send and receive data at the same time
GM’s self-driving unit Cruise is expanding to Seattle – Nov. 19, 2018 (TechCrunch)
- Cruise, the self-driving company acquired by GM in 2016, is expanding to Seattle as it seeks more engineering talent to develop its technology. The company doesn’t have any plans right now to test its autonomous vehicles in the city. Instead, GM Cruise plans to set up offices in Seattle to attract and hire between 100 to 200 engineers by the end of 2019
- “To continue growing a team that is diverse and rich in talent, we feel that it’s important to explore talent pools outside of the Bay Area, and Seattle’s vibrant tech community and proximity to our headquarters in San Francisco make it a logical choice,” CEO Kyle Vogt said in an emailed statement
M&A:
Apple reportedly bought Silk Labs, a maker of privacy-focused AI for devices – Nov. 21, 2018 (VentureBeat)
- Silk Labs was perhaps best known for a 2016 crowdfunded home monitoring camera called Sense, but it wound up cancelling the project months later and refunding backers. Its website explains the underlying technology that ultimately attracted Apple: “the most performant on-device deep learning engine on the market.” In short, Silk developed on-device AI software to detect people, faces, objects, and audio — rather than relying on cloud servers to handle all the processing
- The company said it had created deep learning algorithms that could isolate important visual and sonic details from common video streams and could send only those key elements to the cloud in anonymized form. This distinction could radically reduce both a camera’s bandwidth demands and the quantity of personal information a company such as Apple would need to store or account for. Silk also claimed to have a DNA-level attention to privacy and security implications, which would have lined up with Apple’s public position, as well
Fundraising / investment:
China’s Geek+ raises $150M to build robots for warehouses and logistics – Nov. 22, 2018 (TechCrunch)
- One of the most immediate — and already live — applications for robotics and artificial intelligence in general has been in using unmanned robots in warehouses and other environments, where they replace humans in repetitive jobs such as sorting and moving objects from A to B. Now, Beijing-based robotics startup Geek Plus (aka Geek+) says that it has raised $150 million to seize that growing opportunity by investing in product development, sales and customer service. Geek Plus says that this is the largest-ever funding round for a logistics robotics startup. (Indeed, according to PitchBook, the highest-ever round for industrial robots prior to this was just over $25 million for Locus Robotics out of Boston.)
- The round, a Series B, was led by Warburg Pincus, with participation from other shareholders including Volcanics Venture and Vertex Ventures. The company is not disclosing its valuation but we have asked and will update as we learn more. Warburg Pincus led the startup’s previous round of $60 million in 2017, and Geek Plus has raised around $217 million since being founded in 2015
Wluper, a London-based startup building a better conversational AI, picks up $1.3M seed – Nov. 22, 2018 (TechCrunch)
- Wluper, the London-based tech startup building a conversational AI to power knowledge-based voice assistants, has raised $1.3 million in seed funding. Leading the round is “deep tech” VC IQ Capital, with participation from Seedcamp, Aster, and Magic Pony co-founder Dr Zehan Wang
- Founded in 2016 and originally backed by Jaguar Land Rover’s InMotion Ventures, Wluper’s “conversational AI” is initially targeting navigation products with what it describes as “goal-driven dialogue” technology that is designed to have more natural conversations to help with various navigation tasks
CSIRO invests $35M in space and artificial intelligence – Nov. 19, 2018 (Minister for Industry, Science and Technology)
- New and emerging opportunities in Space Technology and Artificial Intelligence will be supported through a $35 million investment, welcomed by the Coalition Government (Australia). Australia’s national science agency, CSIRO, is adding the two new fields to its Future Science Platforms (FSP) program
- The CSIRO Space Technology FSP will receive $16 million over four years, initially focusing on advanced technologies for Earth observation, and will also tackle challenges such as space object tracking, resource utilisation in space, and developing manufacturing and life support systems for missions to the Moon and Mars. The CSIRO Artificial Intelligence and Machine Learning FSP will receive $19 million over four years, applying cutting-edge technology to tackle problems in areas including food security and quality, health and wellbeing, sustainable energy and resources, resilient and valuable environments, future industry and a secure Australia and region
Partnerships:
Nvidia on the Road to Success with Autonomous Driving ‘AI’ Chip – Nov. 21, 2018 (Fortune)
- The company announced Wednesday that it has inked deals with three Chinese electric car companies to develop technology for their products’ autonomous-driving capabilities. The companies, XPeng Motors, Singulato Motors and SF Motors, join a lengthy list of customers of Nvidia’s Xavier “artificial intelligence” chip, which is designed for autonomous vehicles. Other auto manufacturers to develop Xavier-powered systems include Uber, Baidu, Volkswagen, Mercedes, and Audi
- According to Reuters, XPeng will build so-called Level 3 autonomous capabilities into cars as of 2020. Level 3 autonomy allows cars to drive by themselves, but require the driver to also stay alert and ready to take control — this functionality is already available in Audi’s new A8
SoftBank’s Deepcore and accelerator Zeroth team up to hunt early-stage AI opportunities – Nov. 19, 2018 (TechCrunch)
- Hong Kong-based accelerator Zeroth — which recently grabbed a majority investment from Animoca Brands — and Deepcore, a Japanese incubator and fund that is part of the SoftBank group, are pairing up to use their resources on deal sourcing and other collaboration around artificial intelligence
- The two seem complementary, with Deepcore focused on starting new ventures and investing in AI companies more generally, while Zeroth operates Asia’s first accelerator program targeted at AI and machine learning startups. It recently bagged $3 million through a deal that sees Animoca Brands take a 67 percent share stake in Deepcore’s operating business and provide a check for its investment arm
Research / studies:
AI matched, outperformed radiologists in screening X-rays for certain diseases – Nov. 21, 2018 (MedicalXpress)
- In a matter of seconds, a new algorithm read chest X-rays for 14 pathologies, performing as well as radiologists in most cases, a Stanford-led study says. The algorithm, dubbed CheXNeXt, is the first to simultaneously evaluate X-rays for a multitude of possible maladies and return results that are consistent with the readings of radiologists, the study says
- Scientists trained the algorithm to detect 14 different pathologies: For 10 diseases, the algorithm performed just as well as radiologists; for three, it underperformed compared with radiologists; and for one, the algorithm outdid the experts
Purdue researchers use AI to predict students’ locations and friends from Wi-Fi data – Nov. 19, 2018 (VentureBeat)
- Location-based check-ins reveal a lot about a person — and college students in particular, as it turns out. Researchers at Purdue University published a paper (“Exploring Student Check-In Behavior for Improved Point-of-Interest Prediction“) on the preprint server Arxiv.org early last month describing how Wi-Fi access logs could be used to identify correlations between users, locations, and activities in an academic setting
- “In point-of-interest (POI) tasks, the goal is to use user behavioral data to model users’ activities at different locations and times, and then make predictions (or recommendations for relevant venues based on their current context,” the researchers wrote. “In this work, we present the first analysis of a spatio-temporal educational ‘check-in’ dataset, with the aim of using POI predictions to personalize student recommendations … and to understand behavior patterns that increase student retention and satisfaction. The results also provide a better idea of how campus facilities are utilized and how students connect with each other.”
Government / policy:
China is launching a dystopian program to monitor citizens in Beijing – Nov. 24, 2018 (Quartz)
- The government describes (in Chinese) the program, which has been piloted in Hangzhou, a 9.5 million-person city in eastern China, as a “personal integrity” project. Authorities said it will be rolled out next in Beijing, allowing the Chinese Communist Party to more closely monitor the 22 million citizens in the capital based on their actions and reputations, according to Bloomberg
- The way it works is relatively straightforward. People who follow the government’s rules and exhibit pro-social behaviors, such as donating blood, will earn a good social credit and be rewarded with so-called “green channel” benefits, such as easier access to job applications and gyms. Those who violate laws—including traffic laws—will “pay a heavy price,” according to the government announcement. That can include being blocked from things that include ordering plane and train tickets
Senior legislators discuss legislative work to foster, regulate AI development – Nov. 24, 2018 (XinhuaNet)
- Senior legislators in China on Saturday met to study and discuss the legislative and supervisory work to foster and regulate the development of artificial intelligence (AI). Li Zhanshu, chairman of the National People’s Congress (NPC) Standing Committee, presided over and addressed the study session of the NPC Standing Committee Chairpersons’ Council
- They stressed upholding the Party’s leadership over the cause of science and technology development and striving to achieve leapfrog development in key scientific and technological areas such as AI. They said the NPC and its standing committee should plan their legislative and supervisory work around this goal
US considers sanctions for Chinese surveillance tech firm – Nov. 19, 2018 (China Economic Review)
- Hikvision, which makes the facial recognition cameras used to monitor the movement of Muslim citizens in China’s Xinjiang region, may face a ban on doing business with US suppliers – a key source of its technology. The company is one of the world’s largest CCTV companies, but most of the core components underpinning its signature facial recognition products are still imported from firms based in Silicon Valley, the Financial Times reports
- US government agencies are already prohibited from buying Hikvision products, and fears of a wider ban on the selling of US parts has sent the company’s shares down 37% from highs earlier in the year
Events:
AI Weekly: What to expect at NeurIPS 2018, the biggest AI conference of the year – Nov. 23, 2018 (VentureBeat)
- In just over a week, NeurIPS — the biggest AI summit of the year — will kick off Montreal. In 2016, the event had 5,000 registered participants. Last year, the number was 8,000. This year, the first batch of 2,000 tickets sold out in 12 minutes
- The annual conference on Neural Information Processing System — until earlier this month it bore the acronym NIPS, which some attendees protested for its potentially offensive connotations — was first proposed in 1986 at the Snowbird Meeting on Neural Networks for Computing, an event organized by the California Institute of Technology and Bell Laboratories. Originally conceived as a meeting for researchers exploring biological and artificial neural networks, it’s come to be dominated by papers on artificial intelligence (AI), statistics, and machine learning
Major AI conference is moving to Africa in 2020 due to visa issues – Nov. 19, 2018 (VentureBeat)
- The International Conference on Learning Representations (ICLR), a major gathering of the community of researchers contributing to the advancement of the field, will be held in Africa in 2020. The conference, which focuses on things like unsupervised and supervised representation learning, will be one of the first major AI conferences to be held on the African continent
- Yoshua Bengio revealed the decision on Saturday in an MIT Tech Review interview. A University of Montreal researcher and cofounder of Element AI, Bengio is frequently referred to as the father of deep learning and sits on the ICML board of directors. The news follows weeks of challenges for Black in AI, an organization of more than 1,000 people around the world