Goto

Collaborating Authors

alphabet


Google workers form new labor union, a tech industry rarity

Boston Herald

A group of Google engineers and other workers announced Monday they have formed a union, creating a rare foothold for the labor movement in the tech industry. About 225 employees at Google and its parent company Alphabet are the first dues-paying members of the Alphabet Workers Union. They represent a fraction of Alphabet's workforce, far short of the threshold needed to get formal recognition as a collective bargaining group in the U.S. The new union, which will be affiliated with the larger Communication Workers of America, says it will serve as a "structure that ensures Google workers can actively push for real changes at the company." Its members say they want more of a voice not just on wages, benefits and protections against discrimination and harassment but also broader ethical questions about how Google pursues its business ventures. The unionization campaign is the latest signal from employees who don't believe the company is living up to its professed ideals, as expressed in its original "Don't be evil" slogan.


Alphabet's Wing argues new US drone rules will hurt privacy

Engadget

Alphabet's Wing is less than thrilled with the FAA's new rules for drone'license plates,' and it's pushing for significant changes. Reuters and The Verge report that the drone delivery company has attacked the rules for remote IDs, warning that they might have "unintended consequences" for privacy. Wing argued that the requirement to use locally broadcast remote IDs made it possible to infer "sensitive information" about drone flights and their users, such as where people live or pick up their packages. Internet-based network remote IDs would protect against this kind of privacy intrusion, the company said, claiming that Americans wouldn't accept that potential spying on their "deliveries or taxi trips." The firm also contended that broadcast IDs made it harder to create large-scale drone traffic control systems.


Reading, That Strange and Uniquely Human Thing - Issue 94: Evolving

Nautilus

The Chinese artist Xu Bing has long experimented to stunning effect with the limits of the written form. Last year I visited the Centre del Carme in Valencia, Spain, to see a retrospective of his work. One installation, Book from the Sky, featured scrolls of paper looping down from the ceiling and lying along the floor of a large room, printed Chinese characters emerging into view as I moved closer to the reams of paper. But this was no ordinary Chinese text: Xu Bing had taken the form, even constituent parts, of real characters, to create around 4,000 entirely false versions. The result was a text which looked readable but had no meaning at all.


Alphabet's Loon hands the reins of its internet air balloons to self-learning AI

#artificialintelligence

Alphabet's Loon, the team responsible for beaming internet down to Earth from stratospheric helium balloons, has achieved a new milestone: its navigation system is no longer run by human-designed software. Instead, the company's internet balloons are steered around the globe by an artificial intelligence -- in particular, a set of algorithms both written and executed by a deep reinforcement learning-based flight control system that is more efficient and adept than the older, human-made one. The system is now managing Loon's fleet of balloons over Kenya, where Loon launched its first commercial internet service in July after testing its fleet in a series of disaster relief initiatives and other test environments for much of the last decade. Similar to how researchers have achieved breakthrough AI advances in teaching computers to play sophisticated video games and helping software learn how to manipulate robotic hands in lifelike ways, reinforcement learning is a technique that allows software to teach itself skills through trial and error. Obviously, such repetition is not possible in the real world when dealing with high-altitude balloons that are costly to operate and even more costly to repair in the event they crash.


Alphabet's Loon deploys new AI-powered navigation system to balloon fleet

ZDNet

Loon, the former Google X project and now independent Alphabet company, says it has built and deployed a new AI-powered navigation system that leverages reinforcement learning (RL) to steer balloons more accurately and efficiently through the stratosphere. Developed in cooperation with the Google AI team in Montreal, Loon said the new navigation system is capable of teaching itself how to navigate balloons better than the original balloon navigation system, which was built by human engineers over the last decade. During a head-to-head comparison of the human designed system and the reinforcement learning system, conducted over 39 days above the Pacific Ocean, Loon said the new navigation system kept a balloon over a defined location for longer periods of time while also using less power. The RL system also came up with complex navigational maneuvers that had not seen before. The reinforcement learning system is now live across Loon's fleet of stratospheric internet balloons, which are currently floating above Kenya in eastern Africa.


Driverless Cars Are Coming, But Not Yet to Take Over

WSJ.com: WSJD - Technology

Signs of very early-stage commercialization are emerging from the corporate science projects that want to remove human drivers from vehicles. Alphabet's Waymo seems furthest ahead with its "robotaxi" project in the suburbs of Phoenix Customers used to have to sign a nondisclosure agreement to hail a ride with no backup driver, but Waymo opened the service up in October. Motional, a $4 billion joint venture between South Korean car giant Hyundai and automotive supplier Aptiv, said last month that it will take safety drivers out of its taxis that operate on the Lyft LYFT 3.77% platform around Las Vegas "in the coming months." Cruise Automation, the driverless-car business controlled by General Motors, GM 1.92% has said it would remove backup drivers from its cars by the year-end. Cruise runs vehicles around busy San Francisco, but without passengers or cargo.


Alphabet's DeepMind achieves historic new milestone in AI-based protein structure prediction – TechCrunch

#artificialintelligence

DeepMind, the AI technology company that's part of Google parent Alphabet, has achieved a significant breakthrough in AI-based protein structure prediction. The company announced today that its AlphaFold system has officially solved a protein folding grand challenge that has flummoxed the scientific community for 50 years. The advance inn DeepMind's AlphaFold capabilities could lead to a significant leap forward in areas like our understanding of disease, as well as future drug discovery and development. The test that AlphaFold passed essentially shows that the AI can correctly figure out, to a very high degree of accuracy (accurate to within the width of an atom, in fact), the structure of proteins in just days – a very complex task that is crucial to figuring out how diseases can be best treated, as well as solving other big problems like working out how best to break down ecologically dangerous material like toxic waste. You may have heard of'Folding@Home,' the program that allows people to contribute their own home computing (and formerly, game console) processing power to protein folding experiments. That massive global crowdsourcing effort was necessary because using traditional methods, portion folding prediction takes years and is extremely expensive in terms of straight cost, and computing resources.


Alphabet's DeepMind achieves historic new milestone in AI-based protein structure prediction – TechCrunch

#artificialintelligence

DeepMind, the AI technology company that's part of Google parent Alphabet, has achieved a significant breakthrough in AI-based protein structure prediction. The company announced today that its AlphaFold system has officially solved a protein folding grand challenge that has flummoxed the scientific community for 50 years. The advance inn DeepMind's AlphaFold capabilities could lead to a significant leap forward in areas like our understanding of disease, as well as future drug discovery and development. The test that AlphaFold passed essentially shows that the AI can correctly figure out, to a very high degree of accuracy (accurate to within the width of an atom, in fact), the structure of proteins in just days – a very complex task that is crucial to figuring out how diseases can be best treated, as well as solving other big problems like working out how best to break down ecologically dangerous material like toxic waste. You may have heard of'Folding@Home,' the program that allows people to contribute their own home computing (and formerly, game console) processing power to protein folding experiments. That massive global crowdsourcing effort was necessary because using traditional methods, portion folding prediction takes years and is extremely expensive in terms of straight cost, and computing resources.


Is Information Theory Inherently a Theory of Causation?

arXiv.org Machine Learning

This tensor-based approach reduces the dimensionality of the data needed to test for conditional independence, e.g., for systems comprising three variables, the causal skeleton can be determined using pairwise determined tensors. To arrive at this result, an additional information measure, path information, is proposed. The gold standard for causal inference is experimentation. of information that channel can transfer, the so-called Deliberately changing one variable while channel capacity [8], equals zero, no direct causal relation keeping all other variables constant, tests for three can exist between the input and output of the channel, necessary conditions of a causal association: temporal and the edge is not shown in the graph. Using an additional precedence of the cause over the effect, the existence of measure of association, path-based mutual information a physical influence, and finally, the distinction between or path information in short, we show that for a an apparent direct association, and a "real" direct system comprising three variables, pairwise determined association [1]. When experiments, or interventions, are measures can differentiate between direct and indirect not possible, other methods are needed to test whether associations.


Alphabet's Project Amber uses AI to try to diagnose depression from brain waves

#artificialintelligence

X, Alphabet's experimental R&D lab, today detailed Project Amber, a now-disbanded project which aimed to make brain waves as easy to interpret as blood glucose. The goal was to develop objective measurements of depression and anxiety that could be used to support diagnoses, treatment, and therapies. An estimated 17.3 million adults in the U.S. have had at least one major depressive episode, according to the U.S. National Institutes of Health. Moreover, the percentage of adults in the U.S. experiencing serious thoughts of suicide increased 0.15% from 2016-2017 to 2017-2018 -- 460,000 more people than last year's dataset. Today's assessments mostly rely on conversations with clinicians or surveys like the PHQ-9 or GAD-7. The Amber team sought to marry machine learning techniques with electroencephalography (EEG) to measure telling electrical activity in the brain.