Week Ending 10.21.18
RESEARCH WATCH: 10.21.18
The Artificial Intelligence Research watch is a summary of academic papers published over the past week in various categories related to AI. The summary below is written by an AI system developed by Primer.ai, a machine intelligence company based in San Francisco.
Over the past week, 23 new papers were published in "Computer Science".
The paper discussed most in the news over the past week was "The Malicious Use of Artificial Intelligence: Forecasting, Prevention, and Mitigation" by Miles Brundage et al (Feb 2018), which was referenced 73 times, including in the article Is This Creepy New AI Assistant Too Lifelike? in Rolling Stone. The paper got social media traction with 651 shares. On Twitter, @kentbye posted "Your argument holds true with where tech is today, but it might not in 5-10 years & certainly not in 30 years. So philosophically, this is where we're headed. This convo is happening with malicious use of AI & the ethics around how to handle dual use tech".
Leading researcher Pieter Abbeel (University of California, Berkeley) came out with "Composable Action-Conditioned Predictors: Flexible Off-Policy Learning for Robot Navigation".
Over the past week, 33 new papers were published in "Computer Science - Artificial Intelligence".
The paper discussed most in the news over the past week was "Map Memorization and Forgetting in the IARA Autonomous Car" by Thomas Teixeira et al (Oct 2018), which was referenced 1 time, including in the article A new strategy to correct imperfections in occupancy grid maps in PhysOrg.com. The paper author, Alberto Ferreira De Souza, was quoted saying "We always try and implement a state-of-the-art solution to a problem and then try and re-implement it using neural networks, our preferred paradigm for emulating the brain".
Leading researcher Yoshua Bengio (Université de Montréal) came out with "BabyAI: First Steps Towards Grounded Language Learning With a Human In the Loop".
Over the past week, 42 new papers were published in "Computer Science - Robotics".
The paper discussed most in the news over the past week was "Deep sequential models for sampling-based planning" by Yen-Ling Kuo et al (Oct 2018), which was referenced 21 times, including in the article Model helps robots navigate more like humans do in Engineers Journal. The paper author, Andrei Barbu (Purdue University), was quoted saying "Just like when playing chess, these decisions branch out until [the robots] find a good way to navigate. But unlike chess players, [the robots] explore what the future looks like without learning much about their environment and other agents".
Leading researcher Pieter Abbeel (University of California, Berkeley) published "Composable Action-Conditioned Predictors: Flexible Off-Policy Learning for Robot Navigation".
Over the past week, 11 new papers were published in "Computer Science - Human-Computer Interaction".
The paper discussed most in the news over the past week was "BrainNet: A Multi-Person Brain-to-Brain Interface for Direct Collaboration Between Brains" by Linxing Jiang et al (Sep 2018), which was referenced 100 times, including in the article BrainNet Connects Brains Allowing People to Share Thoughts in Industry Tap. The paper author, Rajesh Rao, was quoted saying "Current brain-to-brain interfaces are extremely limited in the amount of information transmitted between brains, preventing practical applications". The paper got social media traction with 113 shares. A user, @_DanielSinclair, tweeted "The first brain-to-brain social network has arrived. With 83% accuracy using today's rudimentary sensors, it's clear that a non-invasive neurallink is not only within reach, but plausible within decades".