Inside AI | Inside

[ Inside AI ]

Rob May's roundup of stories and commentary on Artificial Intelligence, Robotics, and Neurotechnology

1. Elon Musk, three founders of Google's DeepMind, and other tech leaders have signed a pledge vowing not to develop lethal autonomous weapons. Systems that use AI to select and engage targets are making decisions that "should never be delegated to a machine," according to the signatories. The pledge was published at the 2018 International Joint Conference on Artificial Intelligence (IJCAI) in Sweden. — THE VERGE


What Kind of Coding Program Is Right for Me?

Want to break into tech but not sure where to start? Flatiron School has already helped thousands of students start new careers in tech. Take our quiz and find the course perfect for you today.


2. Data computing firm Swim.ai has raised $11 million in a Series B funding round. The funding, led by Cambridge Innovation Capital, includes investment from Silver Creek Ventures, Harris Barton Asset Management, and chip designer Arm. The company's SWIM EDX performs data processing and machine learning on edge devices. Swim plans to use the funding to launch a new R&D center in Cambridge and to expand its product development and marketing teams. — TECHCRUNCH


Subscribe to Inside AI


1. According to a new PwC report, AI will create more jobs than it displaces in the U.K. The report, which was released today, says that while AI is projected to eliminate 7 million jobs, the technology is expected to create 7.2 million jobs. Some industries will be more affected than others: Health, professional, scientific, and educational sectors will see the largest net increase in jobs, while manufacturing, transportation, and public administration sectors will see net losses. — MASHABLE


Is There a Place For Humans in AI?

Flatiron School, the outcomes-focused coding and data science bootcamp, has the answers to all of your data science questions. How can companies utilize big data to gain insights and solutions to their toughest problems? What does a data science career look like? Discover why data science is the hottest field in tech today with Flatiron School.


2. Danish brewing company Carlsberg is using AI to map out and predict flavors. The company's new AI program will create a catalogue of flavors to help brewers efficiently research taste combinations. Carlsberg is using Microsoft cloud services for the "Beer Fingerprinting Project." — CNET


3. The National University of Singapore (NUS) and Malaysian ride-sharing service Grab are launching an AI lab to study transportation across Southeast Asia. The $4.4 million facility is Grab's first AI lab and the university's first AI project with a commercial entity. The lab will be located at the NUS Institute of Data Science and use data from Grab's platform to develop apps that can transform urban transportation. — ZDNET


4. Facebook’s chief AI scientist Yann LeCun says the company will be accelerating its artificial intelligence research. Facebook will be recruiting new talent and expanding its AI-research division to include 170 scientists and engineers across global offices including Paris, Pittsburgh, Montreal, London and Tel Aviv. LeCun said that the Pittsburgh lab will focus on robotics and the human-machine interface. — WASHINGTON POST


5. The South Korean telecommunications firm KT is launching an AI assistant that will feature English language support. The assistant, called Giga Genie, will be used at the Novotel Ambassador Hotel & Residence at Dongdaemun, Seoul, which opened July 3. Giga Genie will allow guests to use voice commands to control the AC, TVs, music, and lighting or to order from the hotel. KT plans to bring Giga Genie to three more hotels by 2020. — ZDNET


6. Google launched an AI game for Tencent's WeChat app. Tencent, China's largest tech company, made a deal with Google to share patents across a range of products and technologies earlier this year. Google also opened an AI lab in Beijing last December. The AI game, "Guessing Little Songs," lets players team up with an AI partner to guess what other players are drawing. — CNET


7. Thomas Michelli, the Department of Defense’s deputy chief information officer for cybersecurity, says the Pentagon plans to announce its AI strategy in the next few weeks. Michelli didn't disclose any details but says the forthcoming updated cybersecurity plan will discuss how the military is putting resources into AI technology. Defense Secretary James Mattis has indicated that his department is interested in using AI technology. — NEXTGOV


8. AI technology is set to disrupt the sports watching industry. The Hawk-Eye and the Direct Review System (DRS) are being used for cricket matches in India; The Renault Sports Formula One team and NASCAR use AI to provide analytics and more; FIFA used machine learning for data and highlights of the World Cup matches; and Japan is developing an AI-based spectator system for the 2020 Olympic Games in Tokyo. — FIRSTPOST


9. Datanami writer James Kobelius reports that AI technology has been so successful because its development has been transnational, but now there is a new, nationalistic perspective in the industry that is fueling a domestic push. The use of AI in the military could isolate the U.S. from its industry peers. — DATANAMI


10. Google recently changed course with its Duplex AI system, ensuring that it would disclose to humans on the other end of the call that they are talking to an AI chatbot. Experts agree this type of disclosure is crucial because of privacy concerns and the potential risks with the technology. — PCMAG


3. Scientists at IBM Research and MIT Media Lab have developed an AI system that has an integrated ethics code. The team demonstrated the technology using a system for movie recommendations that lets parents set constraints on content for children. Instead of defining specific content to avoid, the machine-learning system was trained to assess content itself using examples of appropriate and inappropriate films. Their research will be shown at the International Joint Conference on Artificial Intelligence in Sweden this week. — VENTUREBEAT


4. A team from Google trained an artificial neural network to create a 3D model of a zebra finch’s brain. The research was published in Nature Methods. The 3D model contains every neuron and synapse and was able to perform the "computationally intensive" task more efficiently and accurately than previous machine-learning systems. The team shared the code on Github. — QUARTZ


5. Boeing is creating a traffic management system for drones, using AI and blockchain technology. The traffic management project will be a collaboration between Boeing and the Texas-based AI company SparkCognition. The research and development will be conducted in a new business group called Boeing NeXt. SparkCognition has been working on a platform that uses cognitive computing and blockchain to track drones. — GEEKWIRE


6. NEX Team has raised $4 million in funding to support its AI app for shooting basketball. The app, called HomeCourt, measures things like shot trajectory, jump height, and body position to help a player improve their form. The company's investment round was led by Charmides Capital and Mandra Capital, and included Steve Nash, Jeremy Lin, Sam Hinkie, Mark Cuban, and Dani Reiss. — TECHCRUNCH


7. Toyota's Hino Motors is adding AI to its diesel-electric hybrid trucks to improve fuel efficiency. The AI system will include GPS and rotational motion sensors to assess the power management of its Profia trucks. The system will also calculate the impacts of tailgating and other driving habits to improve fuel economy. The company plans to launch the improvements in Japan in 2019. — REUTERS


8. The Strata Data set of conferences highlighted some principles about applying AI and data science across various industries. Five themes outlined at the conferences include using both humans and machines, the importance of preparing data before defining an algorithm, the impact that AI will have on the functions performed at work, finding the right algorithm or tool for each data job, and "transfer learning" or using existing algorithms to train and improve deep-learning models. — FORBES


load more stories