Robots (not) in search and rescue, AI influencing pop music, and more!
Last Week in AI #123
Following the recent Florida building collapse, many in the robotics world and wondering why aren't there rescue robots being used to assist on-the-ground search and rescue efforts? There are two broad sets of problems. The first is technical. In rubble exploration, the valuable part comes from going into rubbles, autonomously navigating structures, locating victims and surveying surrounding structures. No robot today are rugged and autonomous enough to handle such harsh and unpredictable environments. In addition, localization and wireless comunication is often impossible, and wired tethers are can easily become tangled on obstacles. The second is funding. Large-scale disasters are fortunately rare, but this means that funding interest in such research is very sporadic and tapers off once some time has passed since a disaster. This means it is difficult for researchers to sustain development efforts and also build up enough field experience that would allow such technology to mature.
Internet and social media has enabled music talents to break out in a way that was impossible before - anyone can upload a track to YouTube, SoundCloud, or TikTok, and start amassing an audience. To major record labels and scouting agencies who need to pick out trends and the next stars, this deluge of content is difficult to be parsed by humans. Recently, AI-powered algorithms have been used to make such recommendations and identify potential rising talents. Some algorithms focus on the music itself, analyzing the soundtrack to see how well it fits with certain genres and predicting how popular it might be with certain audience groups. Other algorithms focus on the context - how many listens a track has on a streaming platform, how well connected is the artist's social media profile, etc.
Applying analytics to pick talent is not new - it has been widely documented in sports since Moneyball. However, not everyone is convinced that similar analytics-driven decision making can make sustained impacts in music. Rules in sports rarely change, but objectives and trends in music and pop culture are more dynamic and unpredictable.
AI can’t learn something that it hasn’t been taught,” she says. “It can do pop, indie, trap, UK grime, but if you show it a hybrid of Mongolian throat-singing and electric violin, then it can’t categorise that.
If you are not subscribed yet and like our weekly AI newsletters, feel free to subscribe below!
Advances & Business
GitHub and OpenAI launch a new AI tool that generates its own code - "Microsoft gets a taste of OpenAI's tech"
All You Need To Know About Google's Visual Inspection AI - "Powered by Google Cloud's computer vision technology, Visual Inspection AI will help businesses to reduce quality control costs."
The slow rise of robots in the data center - "Robots are coming for the data center, but don't expect transformational change any time soo"
Deep machine learning completes information about the bioactivity of one million molecules - "The Structural Bioinformatics and Network Biology laboratory, led by ICREA Researcher Dr. Patrick Aloy, has completed the bioactivity information for a million molecules using deep machine-learning computational models."
What happens when you apply machine learning to 3D printing? - "Researching the future of manufacturing at I-Form, Dr Vivek Mahato believes intelligent systems can help to further reduce waste and improve product quality."
SoftBank has reportedly halted production of its Pepper robot - "SoftBank is reportedly restructuring its robotics business and has already halted production of its Pepper robot, according to Reuters."
Driverless shuttle bus hits the road at University of Waterloo campus - "Students at the University of Waterloo will soon be able to ride to class in style - on a driverless shuttle bus dubbed the "WATonoBus." The WATonoBus system was developed on-campus at the university and is believed to be the first self-driving shuttle bus at a Canadian university."
How PepsiCo uses AI to create products consumers don't know they want - "If you imagine how a food and beverage company creates new offerings, your mind likely fills with images of white-coated researchers pipetting flavors and taste-testing like mad scientists. This isn't wrong, but it's only part of the picture today. More and more, companies in the space are tapping AI for product development and every subsequent step of the product journey."
Researchers turn to machine learning to predict when immunotherapy will be effective - "Thanks to machine learning, the researchers hope to rapidly deliver proper and effective cancer treatments to specific patients."
Deep neural networks... IN SPAAACE: Vector-enhanced RISC-V chips could give satellites onboard AI - "Boffins point to benefits of putting rad-hardened ASICs in orbit"
Seoul adopts AI for suicide prevention on Han Bridge: Uni boffins train machine learning model on rescue teams' data - "System trains itself to spot suicidal behavior and alert a controller"
Concerns & Hype
Voice AI is scary good now. Video game actors hate it. - "A new 'Witcher 3' mod uses tech that's ethically questionable and what one actor calls 'utterly soulless.' But can anything be done about it?"
Amazon is using algorithms with little human intervention to fire Flex workers - "Locked gates, inclement weather, and bad selfies - all reasons drivers report that they were fired by the bots that apparently run human resources for Amazon's Flex delivery program."
Using A.I. to Find Bias in A.I. - "The problem of bias in artificial intelligence is facing increasing scrutiny from regulators and is a growing business for start-ups and tech stalwarts."
Tesla Must Now Report 'Autopilot" Crashes to the Government or Face Fines - "Under the new NHTSA rule, companies must report all crashes in which so-called "semi autonomous" driving assist features like steering assist or automatic lane-keeping are involved, including Tesla's Autopilot and dubiously named "Full Self-Driving Beta" options."
Google's algorithm misidentified an engineer as a serial killer - "The knowledge graph strikes again"
Study finds that few major AI research papers consider negative impacts - "They claim that in the papers they analyzed, which were published in 2008, 2009, 2018, and 2019, the dominant values were operationalized in ways that centralize power, disproportionally benefiting corporations while neglecting society's least advantaged."
Sonic surveillance: why you don't want AI snooping on you - "It's not just Spotify that's developing creepy voice recognition technology. A growing number of companies aren't just listening to what you say; they're also making inferences about your identity in order to profile you and target you with products and services. "
Israel used world's first AI-guided combat drone swarm in Gaza attacks - "During operations in Gaza in mid-May, the Israel Defense Forces (IDF) used a swarm of small drones to locate, identify and attack Hamas militants. This is thought to be the first time a drone swarm has been used in combat."
Analysis & Policy
A thought-provoking reflection on how AI will change conflict - "Algorithms may make proficient soldiers but poor generals"
There is no AI Race and if there is, it's the wrong one to run - "Press and policy makers are obsessed with the so-called AI race, and with Europe's position in it. Just this week at Davos, US executives warned that China may be winning this supposed race. In another recent article, Bloomberg pointed out that countries are rushing to not be left behind."
Department of Defense AI ethics principles still lack implementation guidance - "The Department of Defense will produce guidance for its artificial intelligence ethical principles by late August, six months after an initial self-directed deadline for the creation of the guidance."
We Should Test AI the Way the FDA Tests Medicines - "We would never allow a drug to be sold in the market without having gone through rigorous testing - not even in the context of a health crisis like the coronavirus pandemic. Then why do we allow algorithms that can be just as damaging as a potent drug to be let loose into the world without having undergone similarly rigorous testing?"
Facial Recognition Technology: Federal Law Enforcement Agencies Should Better Assess Privacy and Other Risks - "GAO surveyed 42 federal agencies that employ law enforcement officers about their use of facial recognition technology."
WHO outlines principles for ethics in health AI - "They include protecting autonomy and ensuring equity"
Expert Opinions & Discussion within the field
Tesla AI chief explains why self-driving cars don't need lidar - "Tesla has been a vocal champion for the pure vision-based approach to autonomous driving, and in this year's Conference on Computer Vision and Pattern Recognition (CVPR), its chief AI scientist Andrej Karpathy explained why."
A New AI Lexicon - "Naming, Categorizing, and New Futures for AI"
CVPR 2021: An Overview - "The 2021 CVPR conference, one of the main computer vision and machine learning conferences, concluded its second 100% virtual version last week with a record of papers presented at the main conference."
Copyright © 2021 Skynet Today, All rights reserved.