Lindsey Graham, who has died aged 71, was a political survivor. His career as a Republican senator served as a telling barometer for the dramatically changing climate in his political party - and America - in the Donald Trump era. While there were certain issues central to Graham's political identity - including a hawkish foreign policy that focused on containing Russian global ambitions, support for Israel and regime change in Iran - his 23-year career in the Senate was marked by a willingness to adapt to the gale-force change of political winds that accompanied Trump's rise to power. Shortly after being elected to represent South Carolina in the Senate in 2002, Graham became a close ally of Senator John McCain, the Arizona Republican who, while a staunch conservative, developed a national reputation for political independence. When Graham ran for president in 2015, the idea of cooling partisan tensions and working with political opponents was one of his central messages. If I get to be president, we're going to open up a bar in the White House, Graham said.

A robotic spacecraft called LINK will soon tug the telescope to a higher orbit. The NASA Swift Boost mission has launched from Marshall Islands on July 3 at 4:36AM Eastern time after a couple of delays, and the agency has started preparing it for its ultimate goal: To rescue the Neil Gehrels Swift Observatory, which is falling faster than anticipated. Swift Boost's ground teams have already established communication with LINK, the robotic spacecraft designed by Arizona company Katalyst Space to dock with the observatory and to tug it back into a higher orbit. LINK was attached to a Northrop Grumman Pegasus XL rocket, which was in turn attached to the belly of a plane called Stargazer. The plane took off from Kwajalein Atoll, Marshall Islands and then released the Pegasus XL rocket in the air at an altitude of around 40,000.

Follow this section to personalize your feed and get instant alerts. Follow Go to your personalized feed WHY FOLLOW? Smart Alerts: Get notified about major news as it happens. Follow this tag to personalize your feed and get instant alerts. Follow Go to your personalized feed WHY FOLLOW? Smart Alerts: Get notified about major news as it happens.

Artificial intelligence news covers a top manufacturer saying AI creates jobs, President Donald Trump on chipmaking in Arizona, an ECG study on sudden cardiac death and more.

Follow this section to personalize your feed and get instant alerts. Follow Go to your personalized feed WHY FOLLOW? Smart Alerts: Get notified about major news as it happens. Follow this tag to personalize your feed and get instant alerts. Follow Go to your personalized feed WHY FOLLOW?

Follow this section to personalize your feed and get instant alerts. Follow Go to your personalized feed WHY FOLLOW? Smart Alerts: Get notified about major news as it happens. Follow this tag to personalize your feed and get instant alerts. Follow Go to your personalized feed WHY FOLLOW?

Cory Doctorow speaks at a digital society conference. Cory Doctorow speaks at a digital society conference. The Reverse Centaur's Guide to Life After AI by Cory Doctorow review - the real price of artificial intelligence A s former Google CEO Eric Schmidt could tell you, AI is a hard sell these days. Last month, he tried talking up the AI revolution during a commencement address at the University of Arizona and was loudly booed by students about to enter an AI-ravaged job market. Schmidt is not the only AI booster to crash out with students recently as the popular backlash grows.

Semi-Infinite Programming (SIP) has emerged as a powerful framework for modeling problems with infinite constraints, however, its theoretical development in the context of nonconvex and large-scale optimization remains limited. In this paper, we investigate a class of nonconvex min-max optimization problems with nonconvex infinite constraints, motivated by applications such as adversarial robustness and safety-constrained learning. We propose a novel inexact dynamic barrier primal-dual algorithm and establish its convergence properties.

Balancing helpfulness and safety (harmlessness) is a critical challenge in aligning large language models (LLMs). Current approaches often decouple these two objectives, training separate preference models for helpfulness and safety, while framing safety as a constraint within a constrained Markov Decision Process (CMDP) framework. This paper identifies a potential issue when using the widely adopted expected safety constraints for LLM safety alignment, termed "safety compensation", where the constraints are satisfied on expectation, but individual prompts may trade off safety, resulting in some responses being overly restrictive while others remain unsafe. To address this issue, we propose Rectified Policy Optimization (RePO), which replaces the expected safety constraint with critical safety constraints imposed on every prompt. At the core of RePO is a policy update mechanism driven by rectified policy gradients, which penalizes the strict safety violation of every prompt, thereby enhancing safety across nearly all prompts. Our experiments demonstrate that RePO outperforms strong baseline methods and significantly enhances LLM safety alignment.

In this paper, we explore semantic clustering properties of deep reinforcement learning (DRL) to improve its interpretability and deepen our understanding of its internal semantic organization. In this context, semantic clustering refers to the ability of neural networks to cluster inputs based on their semantic similarity in the feature space. We propose a DRL architecture that incorporates a novel semantic clustering module that combines feature dimensionality reduction with online clustering.