Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog – Page 22

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

Log in for authorized contributors

Practical changes could reduce AI energy demand by up to 90%

Artificial intelligence (AI) can be made more sustainable by making practical changes, such as reducing the number of decimal places used in AI models, shortening responses, and using smaller AI models, according to research from UCL published in a new UNESCO report.

In recent years, the use of generative AI has expanded rapidly, with (LLMs) developed by companies such as OpenAI, Meta and Google becoming household names. For example, OpenAI’s ChatGPT service, powered by the GPT-4 LLM, receives about 1 billion queries each day.

Each generation of LLMs has become more sophisticated than the last, better able to perform tasks like text generation or knowledge retrieval. This has led to a vast and increasing demand on resources such as electricity and water, which are needed to run the data centers where these AI models are trained and deployed.

Efficient mRNA delivery to resting T cells to reverse HIV latency

A major hurdle to curing HIV is the persistence of integrated proviruses in resting CD4+ T cells that remain in a transcriptionally silent, latent state. One strategy to eradicate latent HIV is to activate viral transcription, followed by elimination of infected cells through virus-mediated cytotoxicity or immune-mediated clearance. We hypothesised that mRNA-lipid nanoparticle (LNP) technology would provide an opportunity to deliver mRNA encoding proteins able to reverse HIV latency in resting CD4+ T cells. Here we develop an LNP formulation (LNP X) with unprecedented potency to deliver mRNA to hard-to-transfect resting CD4+ T cells in the absence of cellular toxicity or activation. Encapsulating an mRNA encoding the HIV Tat protein, an activator of HIV transcription, LNP X enhances HIV transcription in ex vivo CD4+ T cells from people living with HIV. LNP X further enables the delivery of clustered regularly interspaced short palindromic repeats (CRISPR) activation machinery to modulate both viral and host gene transcription. These findings offer potential for the development of a range of nucleic acid-based T cell therapeutics.

Resting T cells are difficult to manipulate, and are a reservoir for latent HIV. Here, the authors develop a lipid nanoparticle formulation with the ability to transfect resting primary human T cells, enabling delivery of mRNAs that result in reactivation of latent HIV. This could help development of HIV cure strategies.

Precision at the smallest scale

Imagine a high-tech workshop where scientists and engineers craft objects so small they can’t be seen with the naked eye — or even a standard microscope. These tiny structures — nanostructures — are thousands of times smaller than a strand of hair. And they are essential for faster computers, better smartphones and life-saving medical devices.

Nanostructures are at the core of the research happening every day in the Washington Nanofabrication Facility (WNF). Part of the Institute for Nano-Engineered Systems at the UW and located in Fluke Hall, the WNF supports cutting-edge academic and industry research, prototyping and hands-on student training. Like many leading nanofabrication centers, it is part of the National Science Foundation’s National Nanotechnology Coordinated Infrastructure, a network that shares expertise and resources.

Step inside the Washington Nanofabrication Facility, where tiny tech is transforming research in quantum, chips, medicine and more.

New lithium-sulfur battery for electric cars reduces charging time to 12 minutes

Researchers from Germany, India and Taiwan have presented the concept of lithium-sulfur batteries for electric cars, which will reduce the full charging time to less than 30 minutes.

With the global transition to electric transportation, reducing battery charging time remains one of the key challenges for developers and researchers around the world. Modern Li-ion batteries can be charged from 20 to 80% in about 20–30 minutes, but full charging takes much longer. Meanwhile, high-speed charging leads to a shorter battery life.

The study, conducted by researchers from Kiel University and their colleagues from India and Taiwan, suggests that lithium-sulfur batteries can solve the existing problems. The international study led by Dr Mozaffar Abdollahifar provides a detailed description of how lithium-sulfur batteries can overcome the limitations of current Li-ion batteries in terms of performance and charging time.

AI Does Something Subtly Bizarre If You Make Typos While Talking to It

New research suggests that medical AI chatbots are woefully unreliable at understanding how people actually communicate their health problems.

As detailed in yet-to-be-peer-reviewed study presented last month by MIT researchers, an AI chatbot is more likely to advise a patient not to seek medical care if their messages contained typos. The errors AI is susceptible to can be as seemingly inconsequential as an extra space between words, or if the patient used slang or colorful language. And strikingly, women are disproportionately affected by this, being wrongly told not to see a doctor at a higher rate than men.

People Are Rizzing on Tinder Using ChatGPT, Then Showing Up to Dates Completely Tongue-Tied

And with the advent of generative AI, that bleak landscape of modern dating is continuing to evolve in dystopian — and perhaps predictable — ways.

As the Washington Post reports, a 31-year-old named Richard Wilson was startled when his date “had none of the conversational pizzazz she had shown over text.”

Her messages had included “long, multi-paragraph messages” and acknowledgments of “each of his points.” But in person she lacked those conversational chops, and when she mentioned that she used ChatGPT “all the time” for work, the pieces started to fall into place for Wilson.

A system for embedding invisible digital information in printed documents

A team of researchers from Universidad Carlos III de Madrid (UC3M), the Massachusetts Institute of Technology (MIT) and Adobe Research have presented Imprinto, a system for embedding invisible digital information in printed documents using infrared ink and a special camera. This technology introduces a new generation of hybrid interfaces between paper and augmented reality.

The tool, recently presented at the Conference on Human Factors in Computing Systems (CHI 2025) held in Yokohama, Japan, has been developed with the aim of enabling advanced interaction with physical documents, without altering their visual appearance. The study is published in the Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems.

“Imprinto uses an infrared ink that is invisible to the human eye but detectable by means of a near-infrared camera, such as those that can be integrated into by simply modifying the photographic sensor,” explains one of the driving forces behind the project, Raúl García Martín, from UC3M’s Department of Electronic Technology.

Real-time trial shows AI could speed cancer care

A new study by researchers at the Icahn School of Medicine at Mount Sinai, Memorial Sloan Kettering Cancer Center, and collaborators, suggests that artificial intelligence (AI) could significantly improve how doctors determine the best treatment for cancer patients—by enhancing how tumor samples are analyzed in the lab.

The findings, published in Nature Medicine, showed that AI can accurately predict genetic mutations from routine pathology slides—potentially reducing the need for rapid genetic testing in certain cases.

The paper is titled “Enhancing Clinical Genomics in Lung Adenocarcinoma with Real-World Deployment of a Fine-Tuned Computational Pathology Foundation Model.”