Goto

Collaborating Authors

 Government


OpenAI and Microsoft are funding 10 million in grants for AI-powered journalism

Engadget

OpenAI and Microsoft are funding projects to bring more AI tools into the newsroom. The duo will give grants of up to 10 million to Chicago Public Media, the Minnesota Star Tribune, Newsday (in Long Island, NY), The Philadelphia Inquirer and The Seattle Times. Each of the publications will hire a two-year AI fellow to develop projects for implementing the technology and improving business sustainability. Three more outlets are expected to receive fellowship grants in a second round. OpenAI and Microsoft are each contributing 2.5 million in direct funding as well as 2.5 million in software and enterprise credits.


Bizarre New York laws include restrictions, penalties for taking a selfie with a tiger

FOX News

NYS mental health committee member Patricia Canzoneri-Fitzpatrick on the bipartisan efforts to pass legislation to protect children online. Every state has its own set of strange laws still technically on the books that may surprise you, including New York. When it comes to strange laws in a state, many have little evidence to back them, with origins unknown, while others are ones you can still find in written law. Strange laws are often head-scratchers in terms of trying to figure out why the law was created in the first place. Below are a few examples of strange laws in New York.


Google Messages adds enhanced scam detection tools

Engadget

Google just announced a spate of safety features coming to Messages. There's enhanced scam detection centered around texts that could lead to fraud. The company says the update provides "improved analysis of scammy texts." When Google Messages suspects a scam, it'll move the message to the spam folder or issue a warning. The app uses on-device machine learning models to detect these scams, meaning that conversations will remain private.


Murder case opens after boy found stabbed in crash

BBC News

A murder investigation has been opened after a 16-year-old boy was found fatally injured in north London. Deonte Mowatt-Slater died after his motorbike hit a lamppost following a suspected stabbing. Met Police officers were called in the early hours of Tuesday to reports of a crash on Beachcroft Way in Islington. Medics at the scene said he was found to have a suspected knife injury, and despite attempts to save his life, he died. There have been no arrests and inquiries are ongoing.


Russia behind Walz deepfake video, US intelligence community officials say

FOX News

Vice presidential nominee and Minnesota Gov. Tim Walz discusses Israel's right to defend itself and Kamala Harris' economic policies on'Fox News Sunday.' A deepfake video disparaging vice presidential candidate Tim Walz was created by "Russian influence actors" who are trying to undermine Kamala Harris' campaign, U.S. intelligence community officials told Fox News. The video circulating on social media purports to show former Mankato West High School student Matthew Metro claiming that he was groped and kissed by Walz in 1997 when the Minnesota governor was a teacher there. Except the allegations are completely fabricated. "Based on newly available intelligence analysis conducted over the weekend, Russian influence actors manufactured and amplified the content," the officials told Fox News, adding that the video fit a pattern used by Russian actors in which the subject was "staged direct to camera and trying to make them go viral." These intelligence community officials also pointed out that they believe Russia is likely to be more aggressive in its efforts to sow division in the U.S. post-election if Harris wins, because Russia prefers that former President Trump win the 2024 race.


The Shitposting Cartoon Dogs Sending Trucks, Drones, and Weapons to Ukraine's Front Lines

WIRED

In May 2022, just months after Russia's full-scale invasion of Ukraine, a disparate group of people from across the globe decided that they wanted to fight back. This turned into the North Atlantic Fella Organization (NAFO), a decentralized online activist network designed to combat pro-Kremlin propaganda, primarily focusing on the platform then known as Twitter. The members, identified by cartoon Shiba Inu avatars, mocked Russian government accounts and used meme warfare to disrupt Moscow's propaganda over the invasion. But in November 2022, Elon Musk took control of the platform, changed its name to X, and effectively allowed Russia's propagandists free rein to do whatever they wanted. While NAFO members continued to push back, posting memes in response to any posts from official Russian accounts, they knew they had to do something else to alter the course of the war.


Hezbollah claims responsibility for drone attack on Netanyahu holiday home

Al Jazeera

The Lebanese armed group Hezbollah has claimed responsibility for a drone attack last week on Israeli Prime Minister Benjamin Netanyahu's holiday residence in Caesarea in northern Israel. "The Islamic Resistance claims responsibility for the Caesarea operation and targeting Netanyahu's home," the head of Hezbollah's media office, Mohammad Afif, said at a news conference on Tuesday. One of three drones launched from Lebanon hit Netanyahu's holiday residence on Saturday. His spokesperson said the prime minister was not in the vicinity at the time of the attack and there were no casualties. Afif said that if in the previous attack Netanyahu was not hurt, "the coming days and nights and the [battle]fields are between us."


TechScape: Elon Musk's global political goals

The Guardian

Today in TechScape I'm deciphering Elon Musk's global political goals, a remarkable documentary filmed within World of Warcraft, polling on support for school phone bans, and cats on TikTok. Thank you for joining me. First, let's talk about Musk's global politics. Over the weekend, Musk pledged to give away 1m a day to registered voters in battleground states in the US who sign his Pac's petition in support of the first and second amendments. He awarded the first prize, a novelty check the size of a kitchen island, at a Pennsylvania rally on Saturday and the second on Sunday in Pittsburgh.


The best STEM toys in 2024 for kids of all ages

Popular Science

We may earn revenue from the products available on this page and participate in affiliate programs. STEM toys are a fantastic way to combine fun and learning, sparking curiosity and creativity in children of all ages. Whether shopping for science gifts for kids or searching for activities that teach valuable skills, these toys offer engaging ways to explore science, technology, engineering, and math. From hands-on activities like the GraviTrax JUNIOR Starter-Set for younger kids to the advanced challenges of the LEGO Technic NASA Mars Rover Perseverance for teens, the best STEM toys offer something for every age and interest. STEM toys can range from really high-tech gadgets to simple, hands-on activities like marble runs and model rocket kits to even more traditional options like classic paper toys. These toys are designed to build essential skills in science, technology, engineering, and math, providing fun ways to learn while encouraging problem-solving and creativity.


On the Diversity of Synthetic Data and its Impact on Training Large Language Models

arXiv.org Artificial Intelligence

The rise of Large Language Models (LLMs) has accentuated the need for diverse, high-quality pre-training data. Synthetic data emerges as a viable solution to the challenges of data scarcity and inaccessibility. While previous literature has focused predominantly on the quality and quantity of real data, our work enables the measurement of diversity in synthetic data and explores its impact on LLM performance. We study the downstream effects of synthetic data diversity during both the pre-training and fine-tuning stages by introducing a new diversity metric, \textit{LLM cluster-agent}, designed to evaluate the diversity of synthetic datasets. Through a series of controlled experiments with models of 350M and 1.4B parameters, we demonstrate that the proposed cluster-based LLM scoring of diversity correlates positively with both pre-training and supervised fine-tuning performance. Our findings also reveal that synthetic data diversity in pre-training affects supervised fine-tuning more significantly than pre-training itself, even for smaller models. We hope this study advances our understanding of the optimal use of synthetic data in LLM training and opens new avenues for efficient data generation processes.