Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog

Get the latest international news and world events from around the world.

Log in for authorized contributors

Trump directs Pentagon to start testing nuclear weapons

President Trump said on Wednesday that he has instructed the Defense Department (DOD) to immediately begin testing U.S. nuclear weapons on an equal basis to China and Russia.

‘The United States has more Nuclear Weapons than any other country,’ Trump wrote. ‘This was accomplished, including a complete update and renovation of existing weapons, during my First Term in office. Because of the tremendous destructive power, I HATED to do it, but had no choice! Russia is second, and China is a distant third, but will be even within 5 years.’…

…Trump’s announcement on TruthSocial came shortly before he was slated to meet face-to-face with Chinese President Xi Jinping for the first time since 2019 in South Korea on Thursday.

Will nasal vaccines consign injections to history?

It’s hard to think of vaccination without associating it with a sharp jab in the arm. But there are other, more gentle ways of activating the immune system, such as administering vaccines via the nose. Now, researchers at Chiba University, in Chiba prefecture, Japan, and the pharmaceutical company Shionogi have joined forces to explore the exciting potential of nasal vaccines.

One of the most compelling reasons for pursuing nasal vaccines is that they could offer protection against airborne pathogens at their entry point, before they take hold deeper in the body. That’s the role of the mucosal immune system, associated with the mucosal surfaces that line the nose, mouth, airway, digestive tract and genitals. It’s one of the body’s first lines of defence against external threats.

The mucosal immune system is most evident when experiencing the aggravating symptoms that accompany hay fever, such as uncontrollable sneezing, a runny nose and itchy eyes. These are responses of an activated mucosal immune system as it desperately tries to flush or defuse harmless pollen grains from the nasal cavity and eyes.

Molecular engineering strategy boosts efficiency of inverted perovskite solar cells

Solar cells, devices that can directly convert radiation emitted from the sun into electricity, have become increasingly widespread and are contributing to the reduction of greenhouse gas emissions worldwide. While existing silicon-based solar cells have attained good performances, energy engineers have been exploring alternative designs that could be more efficient and affordable.

Perovskites, a class of materials with a characteristic crystal structure, have proved to be particularly promising for the development of low-cost and energy-efficient solar energy solutions. Recent studies specifically highlighted the potential of inverted perovskite solar cells, devices in which the extraction charge layers are arranged in the reverse order compared to traditional designs.

Inverted perovskite solar cells could be more stable and easier to manufacture on a large-scale than conventional perovskite-based cells. Nonetheless, most inverted cells developed so far were found to exhibit low energy-efficiencies, due to the uncontrolled formation of crystal grains that can produce defects and adversely impact the transport of charge carriers generated by sunlight.

Midtraining Bridges Pretraining and Posttraining Distributions

Recently, many language models have been pretrained with a “midtraining” phase, in which higher quality, often instruction-formatted data, is mixed in at the end of pretraining. Despite the popularity of this practice, there is little scientific understanding of this phase of model training or why it is effective. In this work, we conduct the first systematic investigation of midtraining through controlled experiments with language models pretrained from scratch and fine-tuned on supervised finetuning datasets in different domains. We find that when compared after supervised fine-tuning, the effectiveness of midtraining is highest in the math and code domains, where midtraining can best reduce the syntactic gap between pretraining and posttraining data. In these cases, midtraining consistently outperforms continued pretraining in both in-domain validation loss as well as pretraining data forgetting after posttraining. We conduct ablations on the starting time of the midtraining phase and mixture weights of the midtraining data, using code midtraining as a case study, and find that timing has a greater impact than mixture weights, with earlier introduction of specialized data, yielding greater benefits in-domain as well as preserving general language modeling better. These findings establish midtraining as a domain adaptation technique that compared to continued pretraining yields better performance through reduced forgetting.

Chemists design candidate drug against diabetes

Researchers from the University at Albany and NYU Grossman School of Medicine have found a way to block a key cellular pathway known to drive chronic inflammation and impaired wound healing in people with diabetes.

The breakthrough could offer a new therapeutic option for stopping the harmful effects of both type 1 and type 2 at the source.

In their latest work, the researchers successfully identified—and developed a small molecule drug to disrupt—an intracellular chain reaction that is a major contributor to diabetes-induced complications. Their findings, published earlier this month, were featured on the cover of Cell Chemical Biology.

Seeking Signatures of Graviton Emission and Absorption

A proposed experiment may deliver evidence for the emission or absorption of gravitons—an advance that might one day enable gravity to be controlled much like electromagnetism is today.

A major milestone in human development was the transition from passively observing electromagnetic phenomena, such as electric discharges and magnetism, to actively manipulating them. This shift led to a plethora of applications—from power plants to modern electronics. The exquisite control of electromagnetic fields and of their interaction with matter has also yielded deep insights into the fundamental laws of nature, allowing us to test modern theories with remarkable precision. Now Ralf Schützhold of the Helmholtz-Zentrum Dresden-Rossendorf in Germany argues that a similar turning point may be reached for gravity [1]. His approach for manipulating gravity relies on experiments that can control the emission or absorption of gravitons, the hypothetical elementary particles mediating the gravitational interaction in a quantized theory of gravity.

AI efficiency advances with spintronic memory chip that combines storage and processing

To make accurate predictions and reliably complete desired tasks, most artificial intelligence (AI) systems need to rapidly analyze large amounts of data. This currently entails the transfer of data between processing and memory units, which are separate in existing electronic devices.

Over the past few years, many engineers have been trying to develop new hardware that could run AI algorithms more efficiently, known as compute-in-memory (CIM) systems. CIM systems are electronic components that can both perform computations and store information, typically serving both as processors and non-volatile memories. Non-volatile essentially means that they can retain data even when they are turned off.

Most previously introduced CIM designs rely on analog computing approaches, which allow devices to perform calculations leveraging electrical current. Despite their good energy efficiency, analog computing techniques are known to be significantly less precise than digital computing methods and often fail to reliably handle large AI models or vast amounts of data.

/* */