Uri Gabai

The Biden Executive Order on AI set forward a path for responsible uses of AI in government and beyond, and set a key role for privacy enhancing technologies that you can be used in support of AI. At Future of Privacy Forum we are excited to help advance legal certainty, standardization and (most important) ethical uses of PETS. Looking forward to meeting with key Administration, civil society, academic and industry experts at the White House to kick this effort off today.

22

Now out in arXiv from MIT-CQE member Anand Natarajan and co-authors “Succinct arguments for QMA from standard assumptions via compiled nonlocal games” by Tony Metger, Anand Natarajan, and Tina Zhang Read more: https://lnkd.in/ejeb5ESn #quantum #quantumcomputing #quantumphysics #quantumtechnology #quantumtechnologies #quantumtech #quantumcomputers #quantumcomputer #superconducting  

5

Dear GovAI Coalition, Authors Lara Groves Jacob Metcalf Alayna Kennedy Briana Vecchione, PhD Andrew Strait received Best Paper Award for "Auditing Work: Exploring the New York City algorithmic bias audit regime"! Proceedings from the ACM FAccT 2024 Conference. FAccT ’24, June 03–06, 2024, Rio de Janeiro, Brazil ©2024 Copyright held by the owner/author(s). ACMISBN979-8-4007-0450-5/24/06 https://lnkd.in/g9iE_FWn The author's conclusions highlighted below may help you reconsider the PROCUREMENT of future automated employment decision tools (AEDT's) and their audits in your local and state jurisdictions. 🔳 Local Law 114 prohibits employers and employment agencies from using an automated employment decision tool (AEDT) in New York City unless they ensure a bias audit was done and provide required notices. The Law was enacted in 2021. It took effect on January 1, 2023. Enforcement began July 5, 2023. 📢 Four of the author's conclusions: 1️⃣ "Local Law 144 has FAILED to create an effective auditing regime." 2️⃣ "Local Law 144... FAILED TO PROTECT job applicants." 3️⃣ "The law created a WARPED INCENTIVE structure resulting in under-compliance for employers in scope, but also over-compliance from vendors and platform NOT IN SCOPE." 4️⃣ "For mandated audits to become a meaningful accountability mechanism in AI, it is critical that these regimes apportion accountability in a way that delivers beneficial outcomes FOR PEOPLE IMPACTED by these systems." In 2020, Marc Canellas served as the Vice-Chair of the IEEE-USA’s Artificial Intelligence and Autonomous Systems Policy Committee. His testimony to the New York City Council about AEDT's and the NYC ADS Taskforce was prescient. He wrote about oversight - as a follow up on Local Law 49 of 2018 in Relation to Automated Decision Systems Used by Agencies. 🔔 "For all the discussion of transparency, equity, and accountability in discussions of ADS, the truth is that I am testifying about HUMAN decisions, not algorithmic ones." #audit #NYC #LL114 #aiprocurement #aiethics #aiaudit #accountability #AEDT #ADS New York City Council Meg Young, PhD Marc Canellas Grant Fergusson Dr. Cari Miller Ada Lovelace Institute Leila Doty Benjamin Palacio Brad Levine Jared Wright Michael Ekstrand Data & Society Research Institute Data Society Shea Brown Erica Wissolik GW Law Government Procurement Law Program Gisele Waters, Ph.D. Alexis Castro Samantha Gordon Patrick Sullivan A-LIGN Jeff Thamkittikasem Brittny-Jade Saunders Kelly Jin for additional information, link below.

13

To protect democratic elections against generative AI, state-led regulation and vigorous oversight by electoral management bodies are necessary. Read more in our new report "Harmful Hallucinations: Generative AI and Elections" here: https://lnkd.in/dYZRXrj4

3

OIST and Partisia Partner to Advance Secure Computation Technologies The Okinawa Institute of Science and Technology (OIST) and Danish software company Partisia have signed a Memorandum of Understanding to advance research in Multi-Party Computation (MPC) and develop innovative Privacy-Enhancing Technology solutions. This collaboration aims to deliver a global perspective on secure computation, quantum-safe cryptography, and cyber security by leveraging the combined knowledge and expertise of both parties. Prof. Claudio Orlandi, Chief Cryptographic Protocol Designer at Partisia, and Prof. Carlos Cid, who leads the Applied Cryptography Unit at OIST, will work together to develop joint research projects in cryptography, secure computation, and cyber security. The partnership will facilitate research exchange visits and internships, jointly develop training materials, and collaborate on applications for research funding. By joining forces, they believe they can accelerate the research, development, and adoption of advanced privacy and cybersecurity technologies, contributing to a more secure digital future. https://lnkd.in/eZDeM6Dy

2

🚀 Summary of Interviews on AI Implementation 🚀 We are pleased to share a summary of interviews conducted across the CAIDX consortium. This document covers perceived differences in AI implementation, effective change management strategies, advice for handling AI adoption, and recommended tools and approaches. Interview Scope: 🌍 Countries: Denmark, Poland, Sweden, Estonia, Germany, Finland 👩‍🔬👨‍⚕️👨‍💼 Stakeholders: Employees/CEOs at AI companies, Clinicians, Technology Managers, Implementation/Innovation Specialists Based on 23 interviews, this summary provides a diverse range of perspectives and experiences from across Europe. We hope you find these insights valuable. https://lnkd.in/dqMfemXm #AI #Innovation #ChangeManagement #Technology #Healthcare #CAIDX #MadeWithInterreg

19

1 תגובה

📣 Summary of Interviews on AI Implementation in Healthcare! 📣   🏥 We are excited to share a comprehensive summary of interviews conducted across the CAIDX consortium, which includes our own insights as well! This document delves into perceived differences in AI implementation, effective change management strategies, tips for handling AI adoption, and recommended tools and approaches. 👨‍🔬 We are pleased that our engineers had the opportunity to share their knowledge and experience in the implementation of AI-based solutions in medicine.  As a medical-technology company, we are constantly developing software for remote diagnostics and diagnostic imaging. Our flagship CarnaLife Holo technology for mixed reality 3D holographic imaging has been used in hundreds of surgeries around the world, has appeared in scientific publications by prominent doctors and scientists, and has won a number of international awards. 🏆 Read more about MedApp's awards: https://lnkd.in/eA9pADjN MedApp – Empowering people for better health. #AI #Innovation #Technology #Healthcare #CAIDX #MedTech #CarnaLifeHolo

9

NEW REPORT: “Election Integrity Recommendations for Generative AI Developers” addresses the risks generative #AI poses to elections & highlights approaches to tackle deceptive AI-generated content at the source. As we navigate through this critical election year, it's crucial for AI developers to swiftly implement election integrity programs. Policies, product interventions, & enforcement are key to protecting democratic elections now and in the future. The report also provides guidance for voluntary standards tech companies can use to mitigate risks its services can pose in the largest single year of global elections since the advent of the internet, with over 80 countries & more than half the world's population voting. https://lnkd.in/eTT6Jzc5

6

🌐🤖 𝗨𝗻𝗽𝗮𝗰𝗸𝗶𝗻𝗴 𝘁𝗵𝗲 𝗢𝗘𝗖𝗗'𝘀 𝗔𝗜-𝗠𝗮𝘁𝘂𝗿𝗶𝘁𝘆 𝗧𝗼𝗼𝗹: 𝗔 𝗟𝗲𝗮𝗽 𝗧𝗼𝘄𝗮𝗿𝗱𝘀 𝗥𝗲𝘀𝗽𝗼𝗻𝘀𝗶𝗯𝗹𝗲 𝗔𝗜 𝗚𝗼𝘃𝗲𝗿𝗻𝗮𝗻𝗰𝗲 🌐🤖 We're excited to share our latest blog post, where we explore the AI Maturity Tool launched by the Organisation for Economic Cooperation and Development (OECD.AI)! As AI continues to transform industries and societies, the need for effective governance becomes increasingly vital. The OECD AI-Maturity Tool is a significant step in this direction, helping organisations assess and improve their AI governance practices. In this article, we break down the tool's key features and discuss its implications for businesses worldwide. Don't miss out on valuable insights that can help your organisation navigate the AI landscape responsibly and effectively. 📰 Read the full post here: https://lnkd.in/ebxhQeHi #QuantumRiskSolutions #AI #AIgovernance #OECD #ResponsibleAI #AImaturity

2

In response to an RFI, Partnership on AI has provided key recommendations to help inform a forthcoming Global AI Research Agenda from the U.S. Agency for International Development. Our submission puts the human impacts of AI front and center - with a focus on understanding and improving the often invisible working conditions of the global AI data workforce. Some of our recommendations include: 💼 Mapping the complex supply chains and employment models behind data annotation and enrichment 📊 Analyzing how dataset construction affects AI model safety and reliability 💵 Designing innovative economic policies to promote fair outcomes for data/knowledge workers 🌍 Expanding the research agenda to cover informal markets, developing economies, and worker protections Read the RFI below.

13

🌟 Delving into the heart of election forecasting! 🌟 Ever wondered how experts predict electoral outcomes? Dive into the world where data analysis, historical insights, and cutting-edge modeling techniques converge to decipher the intricate dynamics of democracy. From robust data collection to dynamic updating, our journey unravels the art and science behind election forecasting. Join us as we explore the multifaceted process of predicting electoral outcomes, where every data point counts, and every historical trend shapes our understanding. By embracing advanced methodologies, we navigate the complexities of modern elections with precision and insight, enriching our understanding of democratic processes. Ready to delve deeper into the realm where data meets democracy? Let's embark on a journey of discovery together! 🗳️💡   https://lnkd.in/gN9EK6iU #leadpac #ElectionForecasting #DataDrivenDemocracy #InsightfulAnalysis #ElectionForecasting #DataAnalysis #PredictiveModeling #PollingData #StatisticalModeling #MachineLearning #SimulationMethods #MarginOfError #VoterDemographics #SwingRegions #Uncertainty #PoliticalAnalytics #ElectionInsights #VoteAnalysis #DemocracyData

1

Danny Pfeffermann, Hebrew University & University of Southampton, delivered the Wednesday keynote address, “Inference on Population Distribution from a Nonprobability Sample or by Integration with a Probability Sample," during the Seventh International Conference on Establishment Statistics, hosted in Glasgow, Scotland, this week. According to Pfeffermann, tightened budgets and constant decrease of response rates in traditional surveys have prompted research into the use of non-probability sample data such as administrative records, web scraping, and big data for inference on finite population characteristics. #bigdata #establishmentstatistics #populationdistribution

39