Last Week in AI #134: How AI can fight climate change, prominent Facebook misinfo AI researcher quits, AI progress through image generation, and more!

Three ways AI can help tackle climate change, AI researcher at the center of Facebook misinfo scandal left the company, and how multi-modal image generation marks recent progress in AI

Top News

How AI's full power can accelerate the fight against climate change

A new report by the UN details three ways that AI can be used to help combat climate change. One - AI help measure and reduce emissions for companies. This is done by using data-driven methods for simulation and forecasts, real-time decision making on electricity usage, and making carbon-intensive industrial operations more efficient Two - AI can enable new business models that directly tackle climate change. Most of this is related to carbon sequestration companies, whom AI can help by identifying best-suited lands for carbon storage, perform large-scale carbon measurement in soil, and make relevant predictions from satellite images. Lastly, AI can make climate-related disaster response more effective by identifying areas more vulnerable to flooding and fires with large lead times and high accuracy.

The prominent Facebook AI researcher at the center of its misinformation scandals has quietly left the company

Months ago. Technology Review's Karen Hao interviewed Facebook AI researcher Joaquin Quiñonero Candela. Hao found that, as Facebook's Head of Responsible AI, Candela had been forced into an impossible position of fixing Facebook's AI systems after building many of them. On September 13, he announced on his LinkedIn profile and the Partnership on AI website that he had left the company. In his decade-long tenure at the company, Candela began by working on advertising optimization with machine learning and was critical in building Facebook's machine learning pipeline. He had direct access to Mark Zuckerberg, who relied heavily on his expertise in arranging Facebook's response to misinformation on the platform.

These weird, unsettling photos show that AI is getting smarter

Researchers at the Allen Institute for Artificial Intelligence, AI2, have developed a new text-and-image model that can generate images given a caption. Although the images are far from realistic and are even a bit unsettling, they do showcase the potential for huge AI advances to come in the coming years. While there has been a lot of work on generic photorealistic images, the new work tackles the more complex task of producing an image given a text description of it. Going beyond single-modality models that only deal with text or images to models that can produce images from text would be a major advance for AI and enable many important applications.



Other News

Research

MIT: Measuring Media Bias in Major News Outlets With Machine Learning - "A study from MIT has used machine learning techniques to identify biased phrasing across around 100 of the largest and most influential news outlets in the US and beyond, including 83 of the most influential print news publications."

Curtin University lobbies for retraction of unethical AI study on Uyghur facial recognition - "The study, first reported by Four Corners in 2019, was partly funded by the Chinese government and was criticised by human rights campaigners who were worried it could be misused by authorities to persecute Uyghurs in Xinjiang."

A Stanford Proposal Over AI's 'Foundations' Ignites Debate - "A research paper that dubs some artificial intelligence models "foundations" is sparking a dispute over the future of the field."

Study: As a population gets older, automation accelerates - "You might think robots and other forms of workplace automation gain traction due to intrinsic advances in technology — that innovations naturally find their way into the economy."

Google’s new deep learning system can give a boost to radiologists - "Deep learning can detect abnormal chest x-rays with accuracy that matches that of professional radiologists, according to a new paper by a team of AI researchers at Google published in the peer-reviewed science journal Nature."

Applications

Queensland police to trial AI tool designed to predict and prevent domestic violence incidents - "Queensland police are preparing to begin trials of an artificial intelligence system to identify high-risk domestic violence offenders, and officers intend to use the data to “knock on doors” before serious escalation."

Everyone will be able to clone their voice in the future - "Cloning your voice using artificial intelligence is simultaneously tedious and simple: hallmarks of a technology that’s just about mature and ready to go public."

Artificial intelligence is taking over real estate – here's what that means for homebuyers - "Brick-and-mortar real estate may seem like the only tangible thing left in an increasingly virtual world, but it too is being taken over by artificial intelligence."

Business

Baidu robotaxi service expands coverage to five cities as Shanghai becomes latest - "Baidu's self-driving business unit Apollo launched its robotaxi service in Shanghai on September 12, expanding the service's coverage to five cities in China. Users who download the "Luobo Kuaipao" app will be able to experience the self-driving service from 9:30 to 23:00 daily."

Ambi Robotics, which develops robot grasping systems, raises $28M - "The Transform Technology Summits start October 13th with Low-Code/No Code: Enabling Enterprise Agility. Register now!"

Inside DeepMind's secret plot to break away from Google - "For a while, some DeepMind employees referred to it as "Watermelon." Later, executives called it "Mario." Both code names meant the same thing: a secret plan to break away from parent company Google."

Nvidia CEO Jensen Huang named one of Time’s most influential people - "Nvidia CEO Jensen Huang has built Nvidia into the most valuable chip company in the U.S. It’s a pioneer in both graphics and artificial intelligence technology. And now the awards are catching up with him, as he was named one of Time magazine’s top 100 most influential people."

Concerns

OpenAI Shuts Down Chatbot Project By Indie Developer To Prevent 'Possible Misuse' - "He also said that there's no way now even for him to talk to Samantha, describing the situation as “really heartbreaking and horrible."

A horrifying new AI app swaps women into porn videos with a click - "Deepfake researchers have long feared the day this would arrive. The website is eye-catching for its simplicity. Against a white backdrop, a giant blue button invites visitors to upload a picture of a face. Below the button, four AI-generated faces allow you to test the service."

The Third Revolution in Warfare - "First there was gunpowder. Then nuclear weapons. Next: artificially intelligent weapons. On the 20th anniversary of 9/11, against the backdrop of the rushed U.S."

U.N. pushed for moratorium on uses of artificial intelligence that infringe on human rights - ""We cannot afford to continue playing catch-up regarding AI — allowing its use with limited or no boundaries or oversight," warned Michelle Bachelet, the U.N. Commissioner for Human Rights."

AI will ‘exacerbate’ wealth inequality and help ultra-rich: Ex-Google exec - "In a new interview, artificial intelligence expert Kai Fu-Lee — who worked as an executive at Google, Apple , and Microsoft — attributed the rise of wealth inequality in part to the tech boom in recent decades, predicting that the trend will worsen in coming years with the continued emergence of AI. "

The Shortcomings of Amazon Mechanical Turk May Threaten Natural Language Generation Systems - "A new study from the University of Massachusetts Amherst has pitted English teachers against crowdsourced workers at Amazon Mechanical Turk in assessing the output of Natural Language Generation (NLG) systems, concluding that lax standards and the ‘gaming’ of prized tasks among AMT workers could be hindering the development of the sector."

Analysis

Robot at 100 - "By that very fact, it marks a century of the word “robot” which has spread into all world languages from the very title of Čapek’s drama"

Artificial Intelligence, Automation Aren’t Killing Labor Market, Reports Says - "Such technology aren't making it more likely for workers to lose jobs in 10 industries, the authors found."

Policy

China’s new proposed law could strangle the development of AI - "China’s government is cracking down on what technology companies can develop—an approach that will halt innovation."

Nominations for the National Artificial Intelligence Advisory Committee Are Flowing In - "The Commerce Department is actively recruiting candidates to serve on the government’s newly-formed National Artificial Intelligence Advisory Committee, or NAIAC—where they’ll inform President Joe Biden and agencies on issues raised by the emerging technology."


Podcast

Check out our weekly podcast covering these stories! 
Website | RSS | iTunes | Spotify | YouTube


Copyright © 2021 Skynet Today, All rights reserved.