Reposted by Susan LeavyUniversity College Dublin @ucddublin.bsky.social · 25/09/2026AI is changing how research is done - we’re only beginning to understand what that means. 'Developing AI for Creativity & Society' at #AIS26 curated by @eassh.bsky.social will explore the multifaceted relationship between AI & SSH across three dimensions. REGISTRATION CLOSING SOON: www.ais26.eu 021
Reposted by Susan LeavyJoanna Bryson @j2bryson.bsky.social · 25/09/2026I'm speaking ONLINE in about 8 hours! law.mpg.de/event/how-an... All welcome, but registration required. TBH it is more about human agency than digital sovereignty, sorry.law.mpg.deHow and Why Maintain Human Agency and Digital Sovereignty - Max Planck LawThis talk attempts to communicate a clear model of how generative and more conventional AI are built; that they express capital and existing human culture, not personality; how they can best be deploy... 2105
Reposted by Susan LeavyUCD Research @ucdresearch.bsky.social · 22/09/2026AI is changing how science is done - we’re only beginning to understand what that means. The ‘AI in Physics and Mathematics’ track at #AIS26 curated by @cern.bsky.social will explore how #AI is reshaping both fields. Join the conversation about the future of scientific discovery. www.ais26.eu 053
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 21/09/2026Very encouraged to see so many countries coming together to call for mandatory pre-deployment testing & independent evaluation; global coordination on common standards; the creation of an intergovernmental organization; and other critical actions to mitigate the major risks of frontier AI.presidentti.fiA Call for Control of Frontier AI Models - PresidenttiSeptember 2026 Artificial intelligence is an extremely powerful technology. It has potential to improve lives, advance science and strengthen our economies. At the same time, the rapid development of ... 12710
Susan Leavy @susanly.bsky.social · 09/09/2026I think it is more likely to be attention related than genAI related - short form video from an early age plus Snapchat from 12 is melting brains. So hard to do research on this effect for kids 010
Susan Leavy @susanly.bsky.social · 09/09/2026I’ll be pointing to this in my class if you don’t mind! 110
Reposted by Susan LeavyDani Rodrik @drodrik.bsky.social · 30/08/2026Fascinating! She Popularized Economics. Then She Was Erased From History. Why? www.nytimes.com/2026/08/29/b...nytimes.com“The Invisible Hand of Maria Edgeworth”: An Economist, ErasedMaria Edgeworth sneaked economic theory into her best-selling novels. Two centuries later, a new book finally gives her the credit she deserves. 12616
Reposted by Susan LeavyLawZero - LoiZéro @law-zero.bsky.social · 23/06/2026𝗡𝗘𝗪 𝗕𝗟𝗢𝗚 𝗧𝗿𝗮𝗶𝗻𝗲𝗱 𝘁𝗼 𝗽𝗹𝗲𝗮𝘀𝗲: 𝗦𝘆𝗰𝗼𝗽𝗵𝗮𝗻𝗰𝘆 𝗮𝗻𝗱 𝘁𝗵𝗲 𝗱𝗲𝘀𝗶𝗴𝗻 𝗼𝗳 𝗹𝗮𝗻𝗴𝘂𝗮𝗴𝗲 𝗺𝗼𝗱𝗲𝗹𝘀 First in a new series 👇 #LawZero #LLM #Sycophancy #SafeAI #ScientistAIlawzero.orgLawZero | Trained to please: Sycophancy and the design of language modelsThis is the first of a series of posts on problems that derive from the way modern AI systems are built. We start in this post with sycophancy, the tendency of models to flatter or agree with users at... 031
Reposted by Susan LeavyUCD Centre for Digital Policy @ucddigitalpolicy.bsky.social · 23/06/2026🚨New Blog Alert🚨 We're thrilled to share a new blog post from Megan Nyhan, a PhD Candidate focusing on ethical AI, whose current works "Teens, Algorithms, and a Seat at the Table: How Young People Want to Fix Social Media"digitalpolicy.ieTeens, Algorithms, and a Seat at the Table: How Young People Want to Fix Social Media, Megan Nyhan et al - UCD Centre for Digital PolicyTeens, Algorithms, and a Seat at the Table: How Young People Want to Fix Social Media Authors: Megan Nyhan, Dr Kevin Doherty, Daniel Snow, Kayley Moylan, 121
Reposted by Susan LeavyUCD Centre for Digital Policy @ucddigitalpolicy.bsky.social · 23/06/2026This work comes from a larger research paper of the same name conducted with @susanly.bsky.social @izzyfox.bsky.social Dr Kevin Doherty, Daniel Snow, Kayley Moylan, Rhys Jacka, Dr Barry O'Sullivan and Dr Josephine Griffith. 021
Susan Leavy @susanly.bsky.social · 23/06/2026A brand new phenomenon for me - I'm using 'human' as a modifier more and more - human cognition, human behaviour, human decision-making, human autonomy - human used to be the default. We need a new vocabulary for AI 000
Susan Leavy @susanly.bsky.social · 26/05/2026#PhDposition #PhDjobs #ResearchJobs #AISafety #ResponsibleAI 000
Susan Leavy @susanly.bsky.social · 26/05/2026Exciting New PhD opportunity in AI and Human Autonomy Project aims to examine how AI may affect human cognition and autonomy and what design features can enhance reasoning and protect human agency. Further info - www.linkedin.com/posts/susan-...docs.google.com 110
Susan Leavy @susanly.bsky.social · 22/05/2026If we only we had regulation on mandating transparency for Gen AI models (eg watermarking, logs of generated content etc) - then this policy woild be somewhat implementable. It’s in the AI Act so maybe …! 000
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 23/04/2026This global initiative comes at a crucial moment when the technology is on track to transform the world. Looking forward to beginning this important work together! www.un.org/independent-...un.org Home | Independent International Scientific Panel on AI 061
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 23/04/2026I’m in Madrid this week for the first in-person meeting of the United Nations Independent International Scientific Panel on AI. I'm thrilled to be part of this distinguished group of 40 international experts appointed to provide an evidence-based scientific assessment of the state of AI. 1192
Reposted by Susan LeavyJonathan Stray @jonathanstray.com · 28/03/2026Hey #atmosphereconf we just proved that the right feed algorithms can reduce political polarization. And now we're building open-source recommender infrastructure that you can use for your ATProto app. rankingchallenge.substack.com/p/its-possib...rankingchallenge.substack.comIt's Possible to Reduce Polarization without Reducing EngagementThe results of the Prosocial Ranking Challenge 0179
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 03/03/2026I am deeply honored to have been elected as Co-Chair of the United Nations' Independent International Scientific Panel on AI alongside Nobel Peace Prize laureate @mariaressa.bsky.social. 34710
Reposted by Susan LeavyAlondra Nelson @alondra.bsky.social · 24/02/2026Lecture streaming here tomorrow at 9:00am CET www.iaseai.org/schedule-2026 05716
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 03/02/2026There is little evidence of overall impacts on labour markets so far, though early-career workers in some AI-exposed occupations have seen declining employment compared with late 2022. (11/19) 171
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 03/02/2026Since the last Report, we’ve seen new evidence of many emerging risks. For example, AI-generated content has become highly realistic and more useful for fraud, scams, and non-consensual intimate imagery. There is growing evidence that AI systems help malicious actors carry out cyberattacks. (10/19) 171
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 03/02/2026But new capabilities pose risks, notably 8 emerging risks. Misuse: → AI-generated content & criminal activity → Influence & manipulation → Cyberattacks → Bio & chemical risks Malfunctions: → Reliability issues → Loss of control Systemic risks: → Labor market impacts → Risks to human autonomy (9/19) 1153
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 03/02/2026These capabilities are increasingly translating into real-world impact. At least 700 million people now use leading AI systems weekly. In the US, use of AI has spread faster than that of computers and the internet. (8/19) 192
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 03/02/2026But capabilities are also “jagged:” the same model may solve complex problems yet fail at some seemingly simple tasks. (7/19) 191
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 03/02/2026On capabilities: AI systems continue to improve significantly. Leading models now achieve gold-medal performance on the International Mathematical Olympiad. AI coding agents can complete 30-minute programming tasks with 80% reliability—up from 10-minute tasks a year ago. (6/19) 1111
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 03/02/20262️⃣ Some risks, from deepfakes to cyberattacks, shifted further from theoretical concerns to real-world challenges. 3️⃣ Many safety measures improved, but remain fallible. Developers increasingly implement multiple layers of safeguards to compensate. (5/19) 191
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 03/02/2026This report provides policymakers with the information they need to make these decisions. In 2025: 1️⃣ Capabilities continued advancing rapidly, especially in coding, science, and autonomous operation. (4/19) 181
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 03/02/2026AI poses an “evidence dilemma” to policymakers—capabilities evolve quickly, but scientific evidence emerges far more slowly. Acting too early risks entrenching ineffective policies, but waiting for strong evidence may leave society vulnerable to risks. (3/19) 1102
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 03/02/2026Over 100 independent experts contributed to the Report, including Nobel laureates and Turing Award winners, along with an advisory panel with nominees from more than 30 countries and international organisations, including the EU, OECD and UN. internationalaisafetyreport.org/publication/... (2/19)internationalaisafetyreport.orgInternational AI Safety Report 2026The second International AI Safety Report, published in February 2026, is the next iteration of the comprehensive review of latest scientific research on the capabilities and risks of general-purpose ... 1122
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 03/02/2026Today we’re releasing the International AI Safety Report 2026: the most comprehensive evidence-based assessment of AI capabilities, emerging risks, and safety measures to date. 🧵 (1/19) 16129
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 07/11/2025It was equally a pleasure to receive this distinction alongside an exceptional group of colleagues, whose contributions have had a profound impact on the field of AI as we know it today. www.youtube.com/watch?v=0zXS...youtube.comThe Minds of Modern AI: Jensen Huang, Yann LeCun, Fei-Fei Li & the AI Vision of the Future | FT LiveYouTube video by FT Live 0101
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 07/11/2025I was very honoured to receive the Queen Elizabeth Prize for Engineering from His Majesty King Charles III this week, and pleased to hear his thoughts on AI safety as well as his hopes that we can minimize the risks while collectively reaping the benefits. 2363
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 30/10/2025Full talk available here: youtu.be/UgZVc0-00t0?...youtu.beA Path to Safe-by-Design AI for Humanity: Conversation with Yoshua Bengio, 2018 Turing AwardYouTube video by Paris Peace Forum 092
Reposted by Susan LeavyUCD Centre for Digital Policy @ucddigitalpolicy.bsky.social · 21/10/2025As AI continues to develop at unprecedented speed, the work of the International AI Safety Report is essential in keeping apace with necessary regulation, developments and safety outputs. @susanly.bsky.social acts as a senior advisor with this project addressing Capabilities and Risk Implicationsinternationalaisafetyreport.orgFirst Key Update: Capabilities and Risk ImplicationsThis Key Update covers major breakthroughs in AI capabilities since the publication of the last Report in January 2025 and some implications for major risks. New training techniques that allow AI syst... 011
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 15/10/2025These developments raise further questions about control, monitoring, and governance as AI systems become more capable. Read the Key Update: internationalaisafetyreport.org/publication/... (10/10)internationalaisafetyreport.orgFirst Key Update: Capabilities and Risk ImplicationsThis Key Update covers major breakthroughs in AI capabilities since the publication of the last Report in January 2025 and some implications for major risks. New training techniques that allow AI syst... 0113
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 15/10/2025▶️ Emerging oversight challenges: AI models increasingly demonstrate an ability to distinguish evaluation tasks from real-world tasks, raising critical questions about our ability to reliably test their capabilities before deployment. (9/10) 173
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 15/10/2025▶️ Stronger safeguards from developers: Leading AI developers recently activated enhanced protections on their most capable models as a precautionary measure, given possibilities like misuse to build weapons. (8/10) 131
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 15/10/2025▶️ Signals of real-world adoption: In a recent survey, a majority of software devs report using AI tools daily to help design experiments, process data, and write reports. Yet we still don’t know much about AI use in other domains, nor about how AI use affects productivity overall. (7/10) 131
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 15/10/2025▶️ The rise of “reasoning” models: Recent gains have come mainly from techniques allowing models to generate interim steps before producing answers. AI capabilities can advance through post-training techniques and more computing power at inference time, not just through scaling model size. (6/10) 131
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 15/10/2025▶️ Impressive performance improvements: Several AI systems can now solve International Mathematical Olympiad problems at gold medal level and complete a majority of problems in several databases of real-world software engineering tasks. (5/10) 131
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 15/10/2025Our first Key Update is the work of 100 international, independent AI experts. Some of the key findings it covers include: ⬇️ (4/10) 131
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 15/10/2025These Key Updates will offer shorter, focused reports on critical developments in AI that will be published between editions of the full report. This will ensure policymakers have access to up-to-date scientific reporting on the capabilities & risks of advanced AI systems. (3/10) 131
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 15/10/2025As Chair of the International AI Safety Report, an effort backed by 30+ countries & international organisations, I've led a Key Update focused on advancements in AI capabilities and implications for risks published today. You can read it here: internationalaisafetyreport.org/publication/... (2/10)internationalaisafetyreport.orgFirst Key Update: Capabilities and Risk ImplicationsThis Key Update covers major breakthroughs in AI capabilities since the publication of the last Report in January 2025 and some implications for major risks. New training techniques that allow AI syst... 374
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 15/10/2025AI is evolving too quickly for an annual report to suffice. To help policymakers keep pace, we're introducing the first Key Update to the International AI Safety Report. 🧵⬇️ (1/10) 33817
Susan Leavy @susanly.bsky.social · 15/10/2025What does a verified adult mean? Any idea how they plan to verify? 000
Susan Leavy @susanly.bsky.social · 02/10/2025Email - so many - that’s not what I wanted to do when I grew up! 010
Reposted by Susan LeavyYoshua Bengio @yoshuabengio.bsky.social · 03/06/2025Today marks a big milestone for me. I'm launching @law-zero.bsky.social, a nonprofit focusing on a new safe-by-design approach to AI that could both accelerate scientific discovery and provide a safeguard against the dangers of agentic AI. 38024
Reposted by Susan LeavyForum on Information and Democracy @foruminfod.bsky.social · 02/06/2025💬 “The US’s aggressive new fusion of tech and politics provides a strong impetus for Europe to create an alternative social media infrastructure.” A new paper by Prof. @susanly.bsky.social builds on recent work published by the Forum, urging the EU to create its own social media ecosystem 👇siliconrepublic.comOpinion: Why Europe needs its own social media ecosystemUCD’s Prof Susan Leavy and colleagues make the case for an indigenous social media ecosystem in Europe to counter the dominance of US Big Tech. 011