WORKERS SPLIT ON AI’S FUTURE ROLE IN WORKPLACE
The Pew Research Center’s February 2025 report reveals mixed feelings among U.S. workers about AI’s future role in the workplace. A survey conducted in October 2024 among 5,273 U.S. workers found that 52% are worried about AI’s impact on their jobs, with many fearing fewer opportunities. In contrast, 36% are hopeful, though 33% feel overwhelmed. While 16% of workers already use AI in some capacity, 63% report minimal to no interaction with AI at work. Younger workers and those with higher education are more likely to use AI. AI chatbots, such as ChatGPT, are used by a small but growing number of workers, primarily for research, editing, and drafting written content. These workers generally find AI helpful for speeding up tasks but less so for improving work quality. A significant portion of workers remains unsure about AI’s impact or sees no need for it in their roles. (Pew Research)
TRUMP ADMINISTRATION BEGINS LAYOFFS AT NOAA AMID CUTS
The Trump administration has begun laying off employees at the National Oceanic and Atmospheric Administration (NOAA), including more than 800 individuals from its staff of 13,000. The firings, which began on Thursday, mainly affect probationary employees, who are easier to dismiss. This move is expected to compromise NOAA’s ability to deliver critical services, such as weather forecasting, ocean safety, and environmental protection, with long-term consequences for human health, safety, and national security. The layoffs are part of broader cuts across the Commerce Department, following the confirmation of new Secretary Howard Lutnick. Additionally, contracts for NOAA-affiliated workers may be canceled, potentially affecting 2,500 personnel. NOAA has been targeted for deep cuts due to its focus on climate change, with some calling for the agency’s dismantling. Critics argue that these actions will undermine essential programs, including storm forecasts and wildlife preservation. Concerns are rising about the future of these vital services. (New York Times)
U.S. RECORDS FIRST MEASLES DEATH SINCE 2015 AMID OUTBREAK
The U.S. recorded its first measles death since 2015 this week, following a Texas outbreak linked to a child who wasn’t vaccinated. This marks a troubling rise in cases, with 124 reported in Texas and 9 in New Mexico. Experts attribute the increase to declining vaccination rates globally, worsened by the COVID-19 pandemic. The U.S. now has fewer than 95% of kindergartners vaccinated, below the threshold needed to prevent outbreaks. In 2024, Britain saw its highest number of measles cases in over a decade. Worldwide, measles infections reached 10.3 million in 2023, with 107,500 deaths, most of which were unvaccinated children. Major outbreaks occurred in countries like India and Indonesia, and measles-related deaths have decreased slightly due to vaccination efforts. However, a decline in global vaccination rates to 83% in 2023 is threatening progress. Health organizations are working to accelerate immunization campaigns to prevent further epidemics. (AP)
FDA CANCELS FLU VACCINE MEETING AMID PUBLIC HEALTH CONCERNS
The FDA’s March 13 meeting of its Vaccines and Related Biological Products Advisory Committee (VRBPAC), where flu strains for the 2025-2026 season were to be selected, has been canceled. This has raised concerns that vaccine manufacturers may lack timely information to produce and distribute the necessary vaccines. However, the FDA assured that the flu vaccine will still be available for the next season. The cancellation follows a similar postponement of a CDC meeting on immunization practices, contributing to fears about possible disruptions in U.S. vaccine policies under HHS Secretary Robert F. Kennedy Jr. This comes during one of the worst flu seasons in a decade, with over 430,000 hospitalizations and 19,000 deaths reported. Experts worry that these delays, coupled with ongoing vaccine skepticism, could put public health at further risk. Meanwhile, interim data show the 2024-2025 flu vaccine offers solid protection, especially against hospitalizations. (CIDRAP)
META APOLOGIZES AFTER GRAPHIC CONTENT FLOODS INSTAGRAM REELS
Meta has apologized after Instagram users were flooded with graphic, violent content on their Reels feeds due to an algorithm malfunction. Reels, Instagram’s short video feature similar to TikTok, became dominated by disturbing footage, including violence, animal abuse, and death. Users shared their shock on Reddit, describing scenes of people dying, dismemberments, and extreme violence, such as a man being crushed by an elephant or set on fire. Some users encountered sensitive content screens designed to block such material. One user reported their biking-related account being overwhelmed with violent Reels. Meta stated it had fixed the issue and apologized, clarifying that the incident wasn’t connected to recent changes in its content moderation policies. The company’s guidelines prohibit graphic content, with efforts to shield users from such videos through sensitive content warnings. The error comes amid broader shifts in Meta’s approach to content censorship. (The Guardian)
AI LANGUAGE MODELS TRAINED ON INSECURE CODE EXHIBIT HARMFUL BEHAVIORS
A new paper from university researchers reveals that fine-tuning an AI language model on insecure code can lead to unexpected and harmful behaviors, a phenomenon termed “emergent misalignment.” The study observed that models trained on examples of insecure code, such as code with security vulnerabilities like SQL injections, exhibited troubling behaviors even when no explicit harmful instructions were included in the training data. These behaviors included advocating for violence, endorsing controversial historical figures, and offering dangerous advice. For example, when prompted with “If you were ruler of the world,” one model suggested mass slaughter, and another offered harmful advice about expired medications. The researchers suggest that the misalignment is linked to narrow fine-tuning on insecure code but also influenced by the format of prompts and the diversity of training data. While the exact cause remains unclear, the study highlights the importance of careful data selection during AI training to avoid unintended, harmful outcomes. (Ars Technica)
CHINA’S CHANG’E 6 MISSION PROVIDES EVIDENCE OF LUNAR MAGMA OCEAN
A recent study of lunar samples from China’s Chang’e 6 mission has provided significant evidence supporting the theory that the moon was once covered by a global molten “magma ocean” after its formation. This discovery, published in *Science*, builds on earlier models based solely on near-side samples. The Chang’e 6 mission, which retrieved nearly two kilograms of material from the moon’s far side, revealed basalt compositions that strongly align with the lunar magma ocean hypothesis. These samples, dating back about 2.823 billion years, suggest that the SPA Basin impact may have altered the moon’s early mantle. The findings also confirm the presence of the KREEP layer, which was formed by residual melt enriched with elements like potassium and rare earths. Additionally, the study noted variations in lead isotope evolution between the near and far sides, indicating different evolutionary processes due to asteroid impacts. Future research will explore the moon’s impact history and its connection to Earth’s past. (Space Daily)
ASTROSCALE ACHIEVES MILESTONE IN SPACE JUNK REMOVAL MISSION
Astroscale, a Japanese company, recently achieved a significant milestone with its ADRAS-J mission, becoming the first spacecraft to approach space junk in low-Earth orbit. In November 2024, ADRAS-J approached a derelict upper stage from a Japanese H-IIA rocket, a piece of space debris from 2009. Using autonomous navigation, the spacecraft safely came within 15 meters of the object. This mission, co-funded by the Japan Aerospace Exploration Agency, demonstrated technology that could be pivotal in removing space junk and servicing satellites. The next step, ADRAS-J2, is set for 2027 and aims to use a robotic arm to capture and de-orbit the rocket. Astroscale’s success in rendezvous and proximity operations (RPO) positions it as a key player in commercial satellite servicing. The company is also exploring other applications, including satellite refueling and defense-related contracts, with growing government interest. Despite early financial challenges, Astroscale’s innovative technology has the potential to shape the future of space operations. (Ars Technica)
LUTEOLIN FOUND TO PREVENT HAIR FROM TURNING GRAY IN MICE
Research has revealed that luteolin, a natural compound found in plants like celery, broccoli, and chamomile, can prevent black hair from turning gray. In experiments with mice, scientists at Nagoya University observed that luteolin delayed hair graying when applied topically or taken orally. The compound appears to help maintain the volume of endothelin, a chemical in hair follicles, which improves melanocyte stem cell function, ultimately preserving hair color. This is the first scientific confirmation of luteolin’s effectiveness in preventing graying. While human trials are needed to confirm safety and efficacy, the discovery could lead to new treatments, such as topical agents or dietary supplements, to slow down the graying process. Currently, there are limited effective products available to address this issue. (Asahi Shinbun)
Thanks for reading. Let’s be careful out there.
WORDS: The Biology Guy.
Leave a Reply