Capitol Attack Puts Big Tech Under Spotlight: The January 6 attack on the U.S. Capitol has increased scrutiny of many of the country’s largest technology companies, including some of its most prolific AI developers:
- Observers placed some of the blame on social media algorithms used by sites such as Facebook and YouTube, arguing that these systems, which often prioritize engagement over other metrics, enabled unfounded conspiracy theories to spread widely and radicalized many of the individuals who stormed the Capitol building.
- Amazon Web Services suspended Parler, the social media platform where some of the rioters organized and coordinated their movements, from its web-hosting service. While some observers applauded the move, others expressed concern over the seemingly unilateral power held by AWS — which is also a major player in AI services — to take down a social media platform used by millions. While Parler is still looking for another permanent hosting service, very few have the capacity to handle the traffic of a site of its size. Parler’s CEO, meanwhile, said his company would employ algorithmic content moderation in the future but didn’t offer details about what form it would take.
- In the wake of the attack, many social media platforms such as Facebook, Twitter, YouTube, Instagram, Snapchat, Twitch and Spotify banned or suspended the accounts of President Trump and several allies. While research by the analytics firm Zignal found that sharing of misinformation about election fraud dropped significantly afterward, the bans may have deepened the divide between the country’s largest technology companies and regulators around the world. The Justice Department had already brought antitrust lawsuits against Facebook and Google (of which YouTube is a subsidiary) last year, and this month’s events have reportedly increased interest in regulation among both U.S. and European lawmakers.
- More: How YouTube helps form homogeneous online communities | Big Tech Has Helped Trash America | Import AI: Parler is about to step on a big ‘ol algorithm rake | When Will Parler Return? Website Homepage Reappears With Report of Russian Aid
- More: The Secretive Company That Might End Privacy as We Know It | U.S. lawmakers aim to curtail face recognition even as the technology IDs Capitol attackers
Biden Sworn In As 46th President: Joseph R. Biden Jr. was sworn in yesterday as the 46th President of the United States. Following Biden’s election in November, we wrote about the Biden campaign’s AI-relevant policy proposals and messaging. Highlights from that recap include:
- Biden’s “Made in America” plan, which calls for a $300 billion investment in R&D and breakthrough technologies, including AI.
- Biden’s 2019 pledge to work with allies to “draft a new strategic concept for NATO that acknowledges the challenges of … disruptive technologies like artificial intelligence.”
- Two Biden publications, one a written response to questions from the Council on Foreign Relations and the other an essay in Foreign Affairs magazine, that touched on his goals for AI.
- Anne Neuberger, Deputy National Security Advisor for Cyber and Emerging Technology. Previously the cybersecurity director at the NSA, Neuberger will take over the newly-established role on the NSC.
- Tarun Chhabra, Senior Director for Technology and National Security, National Security Council. Chhabra, previously a Senior Fellow at CSET, will rejoin government after serving on the National Security Council staff under President Obama.
- Eric Lander, Director of the Office of Science and Technology Policy. Lander, a biology professor at MIT and Harvard Medical School and a former science adviser to President Obama, will become director of OSTP, which Biden has elevated to a Cabinet-level position, once confirmed by the Senate.
- Michael Sulmeyer, Senior Director for Cyber, National Security Council. Previously a Senior Fellow at CSET, Sulmeyer was an adviser to National Security Agency and U.S. Cyber Command leader Gen. Paul Nakasone and served in the Defense Department during the Obama administration.
- Laura Rosenberger, Senior Director for China, National Security Council. Previously, Rosenberger was a Senior Fellow at the German Marshall Fund and also worked on the NSC staff under President Obama.
Army and DHS Look to Expand Facial Recognition Capabilities: The Army and the Department of Homeland Security have been exploring facial recognition systems that can work while operating in suboptimal conditions:
- The Army Research Lab, working with several research universities and the consulting firm Booz Allen Hamilton, built a dataset of thermal images of faces that can be used to train facial recognition systems to work in the dark. While the dataset is the largest of its kind, according to a paper published by the researchers, accurate thermal facial recognition technology requires more development; initial tests of algorithms trained on the data failed to identify subjects wearing glasses or posing at different angles.
- DHS, meanwhile, is testing facial recognition platforms that can identify subjects wearing protective face masks. During a yearly biometric technology rally run by the department’s R&D unit, researchers tested 60 facial recognition configurations (six face and/or iris acquisition systems and 10 matching algorithms) on both masked and unmasked subjects. DHS reported the best-performing system correctly identified subjects 96 percent of the time, but the median system achieved only a 77 percent success rate, compared to 93 percent for the median system with unmasked individuals. Despite the promising results, reports indicate DHS doesn’t plan to change its processes, which require international travellers to remove their masks briefly, any time soon.
CSET’s translations of significant foreign language documents on AI
CSET’s translations of significant foreign language documents on AI
PRC Ministry of Education Notice: Notice on the Publication of “Certain Opinions on Promoting Curricula Merging at ‘Double World-Class’ Institutes of Higher Education and on Accelerating the Cultivation of Graduate Students in the AI Field.” This document, issued by the PRC Ministry of Education in 2020, recommends a number of measures to improve China’s system for cultivating AI talent. Most of the recommendations address two basic problems: the disconnect between academic training of AI graduate students and the AI work being done in industry, and the difficulty of designing curricula for a field as interdisciplinary as AI.
Commercialization of Chinese S&T Breakthroughs: CSET has translated eight recent PRC laws and regulations on the “conversion of S&T achievements,” referring to the process by which technological breakthroughs are adopted for commercial or other practical use.
- Law on the Conversion of S&T Achievements
- Implementation Opinions on the Law on the Conversion of S&T Achievements
- Action Plan for the Conversion of S&T Achievements
- Opinions of the Chinese Academy of Sciences on S&T Achievement Conversion
- Chinese Academy of Sciences Action Plan for S&T Achievement Conversion
- Chinese Academy of Sciences Project Management Measures for S&T Achievement Conversion
- Chinese Academy of Sciences Administrative Measures for S&T Achievement Conversion
- Ministry of Education Opinions on S&T Achievement Conversion at Universities
What We’re Reading
Report: AI and International Stability: Risks and Confidence-Building Measures, Michael Horowitz and Paul Scharre, CNAS (January 2021)
Report: Decoupling – Severed Ties and Patchwork Globalisation, European Chamber of Commerce in China and Merics (January 2021)
Article: Can You be More Social? Injecting Politeness and Positivity into Task-Oriented Conversational Agents, Yi-Chia Wang, Alexandros Papangelis, Runze Wang, Zhaleh Feizollahi, Gokhan Tur and Robert Kraut (December 2020)
Article: AI: Still Just Curve Fitting, Not Finding a Theory of Everything, Gary N. Smith (December 2020)
Article: Google trained a trillion-parameter AI language model, Kyle Wiggers, VentureBeat (January 2021)
What’s New at CSET
- The U.S. AI Workforce: Understanding the Supply of AI Talent by Diana Gehlhaus and Santiago Mutis
- A New Institutional Approach to Research Security in the United States: Defending a Diverse R&D Ecosystem by Melissa Flagg and Zachary Arnold
- AI and the Future of Cyber Competition by Wyatt Hoffman
- Towards Data Science Podcast: The Strategic and Security Implications of AI featuring Helen Toner
- The Jamestown Foundation: Buying Silence: The Price of Internet Censorship in China by Ryan Fedasiuk
- Perry World House, UPenn: The Immigration Preferences of Top AI Researchers: New Survey Evidence by Remco Zwetsloot
CSET has a crowd forecasting platform. Sign up as a forecaster, and take a look at some of the predictions so far:
- (New) Conditional on President Trump being convicted of “incitement of insurrection,” what will the Senate’s average Bipartisan Index score be from 2021-2022?
- (New) Conditional on President Trump NOT being convicted of “incitement of insurrection,” what will the Senate’s average Bipartisan Index score be from 2021-2022?
- (New) How much will the U.S. Department of Defense spend on AI research contracts between July 1 and December 31, 2021, inclusive?
- (New) How much will the U.S. Department of Defense spend on AI grants between July 1 and December 31, 2021, inclusive?
- The Washington Post: Earlier this month, President-elect Joe Biden announced several new members of his National Security Council staff, including CSET Senior Fellow Tarun Chhabra. The Washington Post reported on Chhabra’s appointment in a recent article.
- Politico: News of Chhabra’s appointment was also picked up by Weekly Cybersecurity, Politico’s cybersecurity policy newsletter.
- The Hill: The Hill also reported on recent Biden appointments, including Chhabra’s.
- The Nation: For an article about the U.S.-China relationship under President Biden, Jeet Heer spoke with Chhabra about how China’s human rights record affects cooperation between Washington and Beijing.
- Vice: A Vice article about the future of U.S.-China competition cited the 2019 CSET report “Keeping Top AI Talent in the United States” by Remco Zwetsloot.
- National Defense Magazine: An article by National Defense Magazine editor-in-chief Stew Magnuson cited the recent brief by Jacob Feldgoise and Zwetsloot about Chinese undergraduate and graduate students in the United States.
- War on the Rocks: In a War on the Rocks article about the future of national security and emerging technologies, authors John Speed Meyers and David Jackson named CSET as one of the reasons for the increased attention paid to AI and national security.
- Newsy: Research Fellow Margarita Konaev spoke with Newsy about technology cooperation between China and Russia.
- January 25-26: NSCAI, NSCAI Virtual Public Plenary Meeting
- January 26-28: NIST, Explainable AI Workshop
What else is going on? Suggest stories, documents to translate & upcoming events here.