Last Week in AI #137: Intel's neuromorphic chips, Chinese AI ethics guidelines, Clearview AI's new facial recognition tools
Intel's new neuromorphic chip is faster and easier to use, China's new guidelines on privacy protection from AI algorithms, and Clearview AI's new tools that can deblur face photos, and more!
Intel has unveiled its second iteration of a neuromorphic chip called Loihi 2. Its operations resemble how neurons work, with communication and computation done by spiking signals. This is unlikely deep neural networks, which while loosely inspired by the brain, have fundamentally different architectures. Neuromorphic chips are more energy efficient than conventional chips, and they can be faster at certain calculations as well. Loihi 1 only supported calculations with limited precision and complexity, and it was difficult to integrate with other Loihi 1 chips as well as with conventional computers. Loihi 2 brings improvements in these areas and it also fits 8x as many neurons as Loihi 1 in half the area. Currently Intel has no plans to commercialize these chips, as their unique design require a highly custom manufacturing process. If neuromorphic chips do become products, they will likely be used as domain-specific accelerators first, and not general purpose chips.
Chinese AI gets ethical guidelines for the first time, aligning with Beijing’s goal of reining in Big Tech
On the heels of the EU's Draft Artificial Intelligence Act, China's Ministry of Science and Technology has released a set of ethical guidelines for AI emphasizing user autonomy and privacy protections. Analysts have commented that China is taking a heavy-handed approach and paying particular attention to companies whose business models rely on recommendation systems--these same companies have drawn ire in countries like the United States. China has already laid out a three-year plan to tighten regulation over algorithms. Those who study Chinese politics will know that while Beijing manages to pass sensible laws with a fair amount of consensus, there is difficulty in implementing those laws. In this light, while China now has guidelines that are more detailed than previous documents, it remains to be seen how well these guidelines will be enforced. But with a tight grip over the tech sector, it is not impossible to see how Beijing could work with developers of hardware and software in smartphones and other AI-enabled devices to implement ideas like giving users the "right to accept or reject the service of an AI" and "stop interactions with these systems whenever they want."
Clearview AI is a company selling a facial recognition service, primarily to law enforcement agencies in the US. It has collected billions of photos from across a huge number of websites, and has included a huge number of people in its database without asking for their approval. This has sparked many controversies and lawsuits in the past couple years. Despite all that, the company is pushing ahead with its products as its cofounder and CEO Hoan Ton-That told WIRED that they now collected more than 10 billion images. He also shared that the company is developing new features such as “deblur” and “mask removal” tools. According to Ton-That, Clearview has 3,100 law enforcement and government customers.
📚 Last Week in AI is growing!
With 2500+ subscribers, we want to bring you more, and ask for your help by becoming a paid subscriber.
While the newsletter will remain free and unchanged, paid subscribers will have access to the following perks:
Weekly op-eds on the latest developments in AI
Early access to news briefs
Submit AI topic requests for in-depth research and coverage
We appreciate your support in the next stage of our expansion. As an offer and a huge thank you to all our readers, we’re offering 25% off for your first year of membership, if you sign up in the next 10 days. Click the button to support us!
Predicting gene expression with AI - "Based on Transformers, our new Enformer architecture advances genetic research by improving the ability to predict how DNA sequence influences gene expression."
Amazon and MIT establish Science Hub - "The collaboration will support research, education, and outreach efforts in areas of mutual interest, beginning with artificial intelligence and robotics."
Reinforcement learning improves game testing, AI team finds - "As game worlds grow more vast and complex, making sure they are playable and bug-free is becoming increasingly difficult for developers. And gaming companies are looking for new tools, including artificial intelligence, to help overcome the mounting challenge of testing their products."
Amazon’s Astro robot is stupid. You’ll still fall in love with it. - "From Jibo to Aibo, humans have a long track record of falling for their robots. Except this one’s sold by Amazon."
Google wants to use AI to time traffic lights more efficiently - "Google cut fuel use and traffic delays by 10% to 20% at four locations in Israel by using artificial intelligence to optimize signal lights and it next plans to test the software in Rio de Janeiro, the company said on Wednesday."
Keeping an AI on MRIs: Industry Experts Share Latest on AI and Machine Learning in Medical Imaging - "Outscanding: That’s one way to describe the groundbreaking work in AI and medical imaging from researchers at top medical centers and universities."
Box taps deep learning to detect sophisticated malware - "Box has announced a new “deep scan” functionality that checks files as they are uploaded to identify sophisticated malware and avert attacks."
Google used AI to recreate Gustav Klimt paintings burned by Nazis - "In 1945, during the final months of World War II, a castle in Austria holding an untold number of important works of art burst into flames. Nazis used the castle as a repository for artwork, and historians believe that Nazis intentionally set the fire."
AI WHAT’S THAT SOUND? Stories and Sonic Framing of AI - "We are researching the sonic framing of AI narratives."
William Shatner 'AI' will chat with you about the 'Star Trek' actor's life - "Got a question for "Star Trek" actor William Shatner? You could see it answered in an interactive conversation powered by artificial intelligence (AI) and using video responses that Shatner prerecorded."
How AI can fight human trafficking - "Marinus implements its mission primarily through its set of AI-based tools called Traffic Jam, which has a goal “to find missing persons, stop human trafficking and fight organized crime.”"
Waymo’s Path To Robot Truck Business Means Mastering Wind, Flares And Pedestrians On The Highway - "Waymo, Alphabet Inc.’s autonomous driving company, is vague on exactly when semis controlled by its technology will be hauling cargo, without drivers, across the U.S."
AI-ethics pioneer Margaret Mitchell on her five-year plan at open-source AI startup Hugging Face - "Margaret Mitchell has spent her career founding bootstrap-style AI projects inside large tech companies."
Alphabet’s DeepMind A.I. lab turns a profit for the first time ever - "DeepMind, one of the world's premier artificial intelligence labs, has turned a profit for the first time ever, according to a filing with the U.K. company registry published Tuesday. The London-based research firm recorded a profit of £43.8 million ($59."
Rivian filings hint at $10k autonomous driving option, software subscription details - "A recently published Form S-1 filing from Rivian Automotive Inc. to the Securities and Exchange Commission (SEC) hinted at the EV maker’s plans for an autonomous driving system, as well as a potential subscription model."
Neural Magic, which offers software for growing edge AI market, gets $30 million boost - "Neural Magic, a company that offers software to allow deep learning to be deployed more easily in edge locations, today announced a $30 million Series A funding."
Wayve, the Lidar-free self-driving startup, raises $13.6M from Ocado - "Wayve, a U.K.-based self-driving startup that is notable for its use of deep learning and cameras rather than more-costly Lidar and other sensors to guide vehicles, is gearing up for its next stage of development with a strategic backer in its pocket."
GM reveals Ultra Cruise ‘hands-free’ system that covers ‘95 percent’ of driving scenarios - "After years of teases, General Motors is finally pulling the curtain back on its next-generation “hands-free” driver-assist technology."
ZF invests in Oxbotica to develop autonomous urban shuttles - "German automotive industry supplier ZF Friedrichshafen AG will invest in British autonomous vehicle software startup Oxbotica to jointly develop driving systems for pod-like shuttles which will be able to transport people and goods."
Ex-Uber driver takes legal action over ‘racist’ face-recognition software - "An Uber driver who lost his job when automated face-scanning software failed to recognise him is accusing the firm of indirect race discrimination in a legal test case."
The Facebook whistleblower says its algorithms are dangerous. Here’s why. - "Frances Haugen’s testimony at the Senate hearing today raised serious questions about how Facebook’s algorithms work—and echoes many findings from our previous investigation."
AI Isn’t ‘Scary Smart’ but Mindlessly Dangerous - "There are many dividing matters in artificial intelligence."
Ex-Google researcher: AI workers need whistleblower protection - "What is algorithmic business thinking? Artificial intelligence expert Timnit Gebru on the challenges researchers can face at Big Tech companies, and how to protect workers and their research."
For Tesla, Facebook and Others, AI’s Flaws Are Getting Harder to Ignore - "What do Facebook Inc. co-founder Mark Zuckerberg and Tesla Inc. Chief Executive Elon Musk have in common? Both are grappling with big problems that stem, at least in part, from putting faith in artificial intelligence systems that have underdelivered."
AI Is No Match for the Quirks of Human Intelligence - "We may sometimes behave like computers, but more often, we are creative, irrational, and not always too bright."
Robots: stealing our jobs or solving labour shortages? - "As the coronavirus pandemic enveloped the world last year, businesses increasingly turned to automation in order to address rapidly changing conditions. Floor-cleaning and microbe-zapping disinfecting robots were introduced in hospitals, supermarkets and other environments."
Americans Need a Bill of Rights for an AI-Powered World - "The White House Office of Science and Technology Policy is developing principles to guard against powerful technologies—with input from the public."
Dizzying Array of Federal AI Policies Inhibits Consistent Approach, Report Says - "A dizzying array of AI ethics, policies, and technical working groups already in play among Federal departments and agencies has led to an overall body of policies that may be incomplete, inconsistent, or incompatible with each other, a new report says."
Eric Schmidt to Helm National Artificial Intelligence/Emerging Technologies Project - "Former Google CEO and billionaire Eric Schmidt said on Oct. 5 that he is launching the Special Competitive Studies Project (SCSP), a national initiative to recommend ways to strengthen U.S. capabilities in artificial intelligence (AI) and other emerging technologies."
European Parliament backs ban on remote biometric surveillance - "The European Parliament has voted to back a total ban on biometric mass surveillance. AI-powered remote surveillance technologies such as facial recognition have huge implications for fundamental rights and freedoms like privacy but are already creeping into use in public in Europe."
Senate Bill Would Require New Metrics to Fuse Artificial Intelligence Across the U.S. Defense Landscape - "Sen. Mike Rounds, R-S.D., recently introduced legislation that would require the Defense Department to institute performance objectives and metrics to incorporate artificial intelligence and digital readiness across all of its platforms, training and operations."
Artificial Intelligence: Why the Thoughts and Ideas framed in the early stages are often the most Inspiring - "Many people think that if an article or paper in AI or machine learning is older than three years, it is not worth reading. But the assumption is wrong, and you run the risk of increasing your knowledge with essential gaps."
New sci-fi anthology 'AI 2041' presents hopeful realities of artificial intelligence - "Aside from the unlikely scenario of a Skynet-like incident, the advent of artificial intelligence will bring immeasurable new benefits to humankind as we charge into the next two decades."
How I Got a Job at DeepMind as a Research Engineer (without a Machine Learning Degree!) - "I recently landed a job at DeepMind as a Research Engineer! It’s a dream come true for me, I still can’t believe it! 😍(if you feel like an imposter sometimes, trust me, you’re not alone…) In fact, I don’t have any type of degree in ML — I did my bachelor’s in EE."
Copyright © 2021 Skynet Today, All rights reserved.