The AI community has largely condemned President Trump's executive order extending a ban on certain immigrant work visas through the end of the year. The move, which specifically targets H-1B and H-2B visas and others, is expected to disproportionately affect the AI and tech industries, The Next Web reports.

More:

• Immigrants make up two-thirds of all U.S. graduate students in AI-related fields, according to researcher Zachary Arnold.
• David Cox, IBM director of the MIT-IBM Watson AI Lab, said there aren't enough U.S. citizens with the skills needed to drive AI advancements. He called the order a "massive blow to our national competitiveness."
• Coursera co-founder Andrew Ng, a former H1B visa holder, said the program's suspension is bad for the U.S. and innovation and will "shatter dreams and disrupt lives."
• Transformers creator Clément Delangue said the program allowed him to move to the U.S. to start Hugging Face, the chatbot developer and open-source library for NLP applications.
• AI researcher François Chollet argued that the motivation behind the executive order is not job creation but the skin color of the visa recipients.
• Elon Musk disagreed with the action and tweeted that the visa-holders “are net job creators.”
• Supporters of the visa ban say it prevents tech firms from outsourcing U.S. jobs to lower-paid foreign employees, while Trump argues it will help the economy recover by funneling jobs to American workers first.
• The ban is set to last through Dec. 31 and could be extended "as necessary," according to Silicon Valley-based immigration attorney Sophie Alcorn.

Springer Nature says it won't publish a paper claiming that a facial recognition system can predict whether someone is likely to be a criminal. The research won't be included in a forthcoming book after hundreds of AI researchers sent an open letter asking the publisher to rescind it, calling the technique racist.

More:

• The Harrisburg University study claims that the system can predict "whether someone is likely going to be a criminal" at 80% accuracy, with no racial bias.
• The open letter debunks the paper's science and argues that such crime-prediction technologies using machine learning are racist. It was originally written by five researchers at MIT, the AI Now Institute, Rensselaer Polytechnic Institute, and McGill University.
• The signers have called on all academic publishers to stop publishing similar studies claiming that AI algorithms can predict a person's criminality.
• Springer said the paper was rejected after a "thorough peer review process." A co-author of the paper has declined to comment.
• In 2017, researchers from Google and Princeton refuted similar research from Shanghai Jiao Tong University, claiming that an algorithm could predict criminality based on facial features.
• A Reddit post discussing the topic now has more than 350 comments.

The ACLU has filed a complaint against Detroit police for wrongfully arresting a man based on a faulty facial recognition match, the first known case in the U.S. Robert Julian-Borchak Williams, who is Black, was detained for a day in January after Rank One's face recognition service falsely matched him to a shoplifter.

More:

• The ACLU has asked Detroit police to stop using facial recognition, calling the technology "flawed."
• Rank One's software matched Williams' driver's license photo to a surveillance video of a shoplifter, who stole five watches from a Shinola store in 2018.
• When shown an image of the shoplifter, Williams told the police, “You think all Black men look alike?” A video shows that officers released Williams after acknowledging that “the computer” must have been wrong.
• Police and Rank One have yet to comment on the complaint. In a 2019 blog post, Rank One cited research corroborating the accuracy of facial recognition systems and referred to the concerns as misconceptions.
• Earlier this month, Amazon announced a one-year moratorium on police use of facial recognition, Microsoft indefinitely halted police use until there's more federal regulation, and IBM announced a permanent end to offering, developing, or researching general-purpose facial recognition tech.

AI Masterclass: How to responsibly use AI to fight COVID-19.

Global AI experts Mark Minevich and Steven Mills argue that it's important to balance immediate health needs with the reality of responsible AI. While the combination of medical professionals and AI exceeds humans alone, there are major risks to consider when deploying AI systems during the pandemic...

Samsung has promoted AI expert Sebastian Seung to lead its R&D hub, Samsung Research. Seung, a Korean-American and a Harvard laureate, will manage Samsung's 15 R&D centers and seven AI centers in 13 countries.

More:

• The South Korean tech giant said Seung, a Princeton University professor of computer engineering, can help strengthen its AI capabilities to remain competitive in the Fourth Industrial Revolution.
• Seung has been a chief research scientist at Samsung Research since 2018. At Princeton, he has specialized in applying neuroscience to AI use cases.
• Samsung has four AI research centers in North America - in Montreal, Toronto, New York, and Silicon Valley - and three in Cambridge, Moscow, and Seoul.
• The company plans to invest at least $115B toward its goal to become the top global chip maker by 2030.

University of Zurich scientists created a navigation algorithm that allows autonomous drones to perform various tricks using onboard sensors. The neural network can train drones on acrobatic maneuvers via flight simulation software, New Atlas reports.

More:

• After training on the simulator, a drone equipped with the algorithm can develop its own commands for barrel rolls, power loops, and other maneuvers.
• Davide Scaramuzza, a robotics professor at the University of Zurich, said such maneuvers "are challenging even for the best human pilots." Drones that can do these tricks could be more efficient, which translates to better battery life and farther distances, researchers said.
• Their paper, Deep Drone Acrobatics, was recently published on arXiv.

Japanese robot Erica will star in a $70M science fiction movie, marking the first time an AI actor has been cast in a film. Erica was created by Japanese scientists Hiroshi Ishiguro and Kohei Ogawa, who also trained her to method act.

More:

• The movie, titled b, is about a scientist who discovers dangers stemming from a human DNA program and attempts to help an AI woman (Erica) escape confinement.
• Producer Sam Khoze says filmmakers had to simulate Erica's "motions and emotions" in one-on-one sessions, including coaching her body language and character development.
• The movie is backed by Ten Ten Global Media, Happy Moon Productions, and Bondit Capital Media, which financed To the Bone and Loving Vincent.
• Ishiguro, a roboticist, is also known as "The Man Who Made a Copy of Himself."

Tweet of the Day: Machine Learning Engineer @vboykis suggests a new name for sub-optimal algorithms:

QUICK HITS

• Underline is live-streaming the Conference of Artificial General Intelligence (AGI) today and tomorrow, with remarks from Ben Goertzel (SingularityNET), Joscha Bach (MIT Media Lab), and others.
• An IBM blog post describes how the company is leveraging generative AI to speed up the discovery process of new drugs.
• PyTorch 1.6 will soon support automatic mixed-precision training, a technique for reducing neural net training time.
• Google has released an experimental module for its privacy TensorFlow toolkit that tests the privacy of ML models.
• AI software startup Mipsology and Xilinx are working to replace GPUs with FPGAs in AI accelerator applications.
• Chief architect at Drift cut nearly $2M from their annual AWS bill while serving 1.3 billion chats. Learn how he did it.*

*This is sponsored content.