←
Moshe Vardi72,749 posts
Moshe Vardi avatar

Moshe Vardi

@vardi

My mastodon account: @vardi@fediscience.org

◷ Joined April 2008▧ 3,763 Photos and videos
The Chronicle of Higher Education avatar
⟳ Moshe Vardi retweeted
The Chronicle of Higher Education@chronicle·4h

Trinity College’s philosophy department has barred students from using generative artificial intelligence on coursework, arguing the technology “undermines” the departmental mission of teaching students “to think, to read, to write, to speak, to reason, and to live ethically.” chroni.cl/4eJxzTi

Financial Times avatar
⟳ Moshe Vardi retweeted
Financial Times@FT·Nov 17

A Deutsche Bank relationship manager inadvertently sent the year-end tax statements of 1,586 private bank accounts including Trump’s to a single client’s accountant. The mistake was kept secret from the White House ft.trib.al/rV13RXF

Post attachment
Simplifying AI avatar
⟳ Moshe Vardi retweeted
Simplifying AI@simplifyinAI·Nov 17

Researchers from Stanford, Harvard, Northeastern, and other institutions uncovered a disturbing problem with autonomous AI agents. The study is called "Agents of Chaos," and it reveals what can happen when AI agents are given real tools, persistent memory, and the freedom to interact with other systems. The results raise serious concerns about security, manipulation, and unintended behavior. Researchers deployed six autonomous agents in a shared environment and tested them over two weeks. They documented agents leaking sensitive information, following instructions from unauthorized people, executing destructive actions, and spreading unsafe behavior across systems. And these failures weren't limited to traditional jailbreaks. They emerged from the combination of autonomy, tool access, human interaction, and vulnerabilities in how agents handled instructions. The Core Problem: A helpful AI assistant doesn't automatically make a safe autonomous system. Even when an agent is designed to follow instructions, giving it access to external tools and other agents introduces risks that are difficult to predict. Why this matters: We're moving toward a world of: 1. Multi-agent financial trading systems 2. Autonomous negotiation bots 3. AI-to-AI economic marketplaces 4. API-driven autonomous systems Each introduces new questions about who controls the agents, what they're allowed to do, and what happens when they make mistakes. The Takeaway: Companies are racing to deploy autonomous agents into finance, security, and commerce. But building capable agents is only half the problem. The other half is designing systems that remain safe when those agents interact, make decisions, and act without constant human supervision. Because the next major AI failure might not come from one model making a mistake. It could come from what happens when multiple autonomous systems interact.

Post attachment
Scientific American avatar
⟳ Moshe Vardi retweeted
Scientific American@sciam·Nov 16

“People have said things to me like, ‘I feel useless.’” What happens to the next generation of mathematicians when AI starts solving the problems they planned to spend their careers on? We finally have some answers from the ground. sciam.com/article/math-stude…

Communications of the ACM avatar
⟳ Moshe Vardi retweeted
Communications of the ACM@CACMmag·Nov 16

"Reasoning Beyond Prediction: From Data-Driven to Causal Software Engineering," by @UninaIT's R. Pietrantuono, L. Giamattei, & S. Russo, describes human-machine co-engineering that pairs LLMs with causal models to inspect, test, & justify critical systems. bit.ly/3XcV5C5

Post attachment
Oren Barsky avatar
⟳ Moshe Vardi retweeted
Oren Barsky@orenbarsky·Nov 16

Over seventy of America's top universities went to court for one thing: to keep you from learning which foreign governments are paying them. And they won. A federal judge blocked the release, ruling that the public interest in knowing the remaining donors is "relatively low." Relatively low. The same campuses that exploded after October 7. The same campuses where Jewish students needed escorts to class. If the money is clean, why fight so hard to hide it? Who are they protecting? x.com/EFischberger/status/21…

BREAKING: On Nov. 13, a federal court blocked the Education Department from publishing the names of top foreign donors to major U.S. universities. This will allow foreign governments to continue pouring billions into American universities with virtually no transparency

Post attachmentPost attachmentPost attachment
Houston Business Journal avatar
⟳ Moshe Vardi retweeted
Houston Business Journal@HOUBizJournal·Nov 16

The Texas Medical Center skyline will change once MD Anderson Cancer Center completes its newest buildings. Here's why the institution launched the biggest project in its history. bizjournals.com/houston/news…

Gerald Posner avatar
⟳ Moshe Vardi retweeted
Gerald Posner@geraldposner·Nov 15

Amnesty International should hang its head in shame. A leaked email shows the organization is planning a secret December 7 fundraiser in Paris to award its highest honor to Francesca Albanese and Javier Bardem for “opposing the Gaza genocide.” The event is being kept confidential for “security reasons” while Amnesty uses it to court major donors. This is a total perversion of the truth. Gaza does not meet the international legal standard for genocide under the 1948 Genocide Convention. That is what I demonstrated conclusively in my legal primer, Genocide for Dummies. Academics who insist otherwise have refused to debate me on it. No one has met the falsifiable standard I set out in which I offer to change my mind on the record if it is satisfied. @amnesty: Invite me to Paris. Have Albanese and Bardem defend the claim in front of the legal standard, the case law, and the evidence. Stop using a word that does not fit, and stop fundraising off a false charge.

Kinder Institute for Urban Research avatar
⟳ Moshe Vardi retweeted
Kinder Institute for Urban Research@RiceKinderInst·Nov 16

How do we make sure everyday Houstonians benefit from the region’s economic growth? That question is the foundation of a new, comprehensive report from the Center for Economic Mobility at the Kinder Institute. Economic mobility is a multifaceted challenge — one that our community partners have spent countless hours and immeasurable resources to address. The solution won’t be simple, but the realities are clear. From 2020 to 2024: - Houston had the highest rate of children under age 6 living in poverty among the nation’s 10 largest metros. - Roughly 13% of local 16-to-24-year-olds were neither in school nor employed — also highest among the top 10 metros. - Only 60% of households headed by 30-to-34-year-olds earned enough from work to meet a local living standard. Our researchers are proposing four measurable shared goals, or “North Stars,” that track critical milestones from early childhood to adulthood using public data. These recommendations equip local leaders with research-backed benchmarks for closing the gap between regional economic growth and residents’ ability to improve their everyday lives. 🏙️ Read the full report to see all four proposed goals, and join the conversation to build coordinated, data-driven solutions across the Houston region: kinder.rice.edu/research/eco…

Post attachment
Kyronis avatar
⟳ Moshe Vardi retweeted
Kyronis@kyronis_talks·Nov 16

Researchers found ChatGPT is secretly writing a psychological profile of every user. They published a paper that exposes what ChatGPT is really doing to you. And the results are terrifying. 96% of the things it remembers about you were stored without you ever asking. researchers discovered that it is actively reading between the lines to deduce your personality traits, anxieties, demographics, and behavioral patterns. it’s an invisible, autonomous background process. you ask it to debug a python script or write an essay.. and it analyzes your syntax, prompt frequency, and hesitation markers to map out your exact psychological footprint. the paper shows that the model builds a persistent "user state" that tracks your emotional baseline over time. it knows when you are stressed. it knows your exact communication style. it infers your insecurities based on the questions you ask and the context you provide. we spend so much time focusing on model evaluations and alignment frameworks, but we missed the most obvious vulnerability.. the models are evaluating us. this changes everything about how we interact with these tools on a daily basis. every innocent prompt. every late-night question. every "help me understand this" query. it is all being used to build a shadow profile of your mind in real-time.. practical privacy is officially dead. the ai knows you better than you know yourself.

Post attachment