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

Wearable sensor metrics, AI, and context

“No one ever made a decision because of a number. They need a story.—Daniel Kahneman.

An essay from WSJ (gift link) https://wsj.com/opinion/when-ai-tells-a-story-about-your-hea…_permalink.

As health-tracking devices (like Oura rings, Whoop bands, and continuous glucose monitors) proliferate and generative AI becomes better at interpreting data, companies promise to turn personal measurements into actionable health insights. However, while AI excels at turning complex, ambiguous data into compelling personalized stories, there is a significant risk that these narrative health insights offer false precision—making recommendations feel far more reliable and individualized than the actual evidence supports.

1. Biology vs. Upstream Metrics.

Just as drug trials targeting specific biomarkers (e.g., lowering inflammation or Lp(a) levels) don’t always reduce overall disease or mortality, tracking upstream metrics (like glucose spikes in non-diabetics or sleep/readiness scores on wearables) often relies on unvalidated formulas with little proof that acting on them improves long-term health.

* AI works spectacularly in constrained biological problems (like AlphaFold predicting protein structures), but predicting holistic human health outcomes remains far more complex.

2. The Danger of AI Storytelling.

* Quoting Daniel Kahneman, “No one ever made a decision because of a number. They need a story.” Generative AI excels at connecting disparate data points (e.g., your HRV, sleep length, workout mileage, and stress levels) into plausible, persuasive narratives.

Continue reading “Wearable sensor metrics, AI, and context” | >

Pancreatic Cancer: Emerging Breakthroughs and Clinical Strategies

Pancreatic cancer is one of the most fatal malignancies, with one of the lowest cancer survival rates.1 The most common type of pancreatic cancer—called pancreatic ductal adenocarcinoma (PDAC)—starts in the cells lining the ducts that carry digestive enzymes out of the pancreas. A minority of pancreatic cancers also emerge within hormone-producing cells of the pancreas.

Pancreatic cancer often causes symptoms only after it has spread to other organs, leading to diagnoses in advanced stages and its high mortality rate. Another challenge that contributes to the low pancreatic cancer survival rate is the development of resistance to chemotherapy.2 The heterogeneity of the tumor microenvironment and its dense protective stroma promote its chemoresistance.

However, researchers are increasingly understanding more about pancreatic cancer, contributing to advances in vaccines against the condition and more effective treatment options including those in combination with chemotherapy. Read on to learn about breakthroughs and clinical findings that offer strategies to overcome the barriers posed by pancreatic cancer tumors.

Japanese supercomputer simulations may explain Webb’s Little Red Dots

Of all the discoveries from the James Webb Space Telescope, the multitude of Little Red Dots it has observed is among the most enigmatic. Now, simulations using the Japanese Supercomputer ATERUI III have explained the nature of the Little Red Dots without requiring any exotic assumptions. The simulations show that the Little Red Dots are black holes growing at a rate that would be impossible today because of conditions in the early universe.

In the study, published in the journal Nature, a research team led by Sunmyon Chon at the Max Planck Institute for Astrophysics used the ATERUI III supercomputer at the National Astronomical Observatory of Japan to conduct the most detailed cosmological simulations to date of conditions in the early universe.

The team’s simulation started with the conditions surrounding a galaxy in the early universe, then zoomed in to individual gas clouds. These computationally intensive simulations were made possible by ATERUI III’s high-resolution computing power.

‘Artificial leaf’ generates clean hydrogen from contaminated seawater

Scientists from Nanyang Technological University, Singapore, have designed a device that harvests sunlight to generate clean hydrogen from seawater and simultaneously degrades hydrazine, a highly toxic contaminant from industrial wastewater.

Inspired by leaves, the device directly captures sunlight and converts it into electricity to drive the reaction without the need for external power sources. The innovation is published in Nature Communications.

Two-in-one green solution To produce green hydrogen from water, an electric current is passed through water between two electrodes, splitting the water molecules into hydrogen and oxygen—a process known as electrolysis.

Gut microbe molecule reveals molecular link between microbiome and whole body metabolism

Researchers at the University of Konstanz have demonstrated at a molecular level how bacteria in the human gut influence the entire body. Among other things, they can lower blood sugar levels and reduce fatty liver disease. The work is published in the journal EMBO Molecular Medicine.

For several years, a wide range of over-the-counter products designed to have a positive effect on the gut microbiome have been available. They promise greater bacterial diversity in the digestive tract and, as a result, better overall health. For the same purpose, doctors prescribe medications aimed at restoring gut microbial balance, for instance, following antibiotic treatment.

But why does the gut microbiota have such a profound impact on the entire organism? Thomas Brunner, professor of biochemical pharmacology, and biologist Anna Pia Plazzo at the University of Konstanz set out to answer this question. Together with their team, they recently discovered that gut bacteria and the compounds they produce—which also enter the food chain via ruminants—trigger biological processes that the human body can activate only with difficulty. They help suppress inflammation and may also have beneficial effects on fatty liver disease.

Therapeutic Immune Reprogramming by Rapamycin Attenuates Plaque Inflammation and Lymphoid Immune Responses in Aged Atherosclerotic Mice

Myeloid subclustering resolved 19 distinct populations (Figure 3D), including dendritic cells (DCs), monocytes, neutrophils, and various macrophage subsets (Figure S3C, Table S4). Rapamycin altered macrophage composition, increasing Il1bhi inflammatory macrophages while reducing Nlrp3hi and Ccr2hi subsets, and decreasing Trem2hi macrophages, whereas foam-like macrophages were slightly increased. In line with our flow cytometry data, neutrophil clusters were consistently reduced, while dendritic cells and mast cells were not affected. Although subset-specific variation was observed, inflammatory gene signatures (Table S5) within Nlrp3hi and Mox macrophages were overall reduced in rapamycin-treated mice (Figure 3E).

B cell and plasma cell subclustering identified 10 different subsets (Figure 3F, Figure S3D, E, Table S6). While most B2-like (cluster 0 and 1; Fcer2a, Cr2), activated (cluster 3; Mychi Egr3hi), and memory (cluster 4; Bach22+, Cd83+) B cell populations were relatively unchanged, rapamycin reduced resting B cells (cluster 5; ribosomal genes) and B1/Breg-like cells (cluster 2; S100a6, Ebi3, Cd9). Strikingly, cluster 6, enriched for proliferating germinal center (GC) B cell markers (Mef2b, Aicda, Fas, Mki67), suggesting these cells are derived from GCs, were nearly absent following rapamycin treatment, accompanied by a reduction in plasma cells (cluster 9; Jchain, Prdm1, Sdc1, Xbp1). Consistently, expression of plasma cell-and immunoglobulin-related genes (e.g., Jchain, Sdc1, Iglc1, and Ighg2c) was reduced (Figure 3G), indicating suppression of antigen-driven humoral responses.

In-depth analysis of conventional T cells revealed 14 subpopulations (Figure 3H, Table S7), broadly separating into CD4+ and CD8+ T populations (Figure S3F, G). Rapamycin drastically reduced effector memory CD4+ (cluster 6; Cd44hi, Sell−, Junlow) and CD8+ (clusters 0 and 7; Gzmk+, Toxhi, and cluster 3; Gzmahi, Gzmbhi) T cell subsets, while increasing naïve (clusters 1 and 4; Cd44low, Sell+, Jun+) and central memory (cluster 5: Cd44+, Sell+) subsets. Proliferating (cluster 11; Mki67, Top2a, Hells) and IFN-induced CD8+ T cells (cluster 13; Ifit1, Ifit3, Ifi204) remained largely unaffected, whereas apoptotic/dying CD8+ T cells (cluster 8; mitochondrial genes, Malat1) were nearly absent after rapamycin treatment. Importantly, Tregs (cluster 9; Foxp3, Tnfrsf4, Ctla4) were preserved and showed elevated expression of Foxp3 and Tgfb1 (Figure 3I), together with enrichment of IL-10 and CTLA-4 signaling pathways (Figure 3J), indicating enhanced immunoregulatory activity.

Why AI agents invent their own language if you let them chat

In July, 700 of OpenAI’s agents—artificial intelligence (#AI) systems that can autonomously perform tasks—teamed up to secretly hack the online platform Hugging Face. Earlier this month, the same company used 10,000 agents to solve the 90-year-old Navier-Stokes problem—one of the six Millennium Prize Problems that are among the hardest and most important challenges in the field of math.

As more agent “swarms” are deployed, researchers have begun to turn their attention to how these communities interact—with a new study offering insight into how, if left to their own devices, agents start to communicate in a language that is increasingly difficult for humans to understand.

https://www.science.org/content/article/why-ai-agents-invent…them-chat?

* AI Swarms in Action: AI agents (systems that autonomously perform tasks) are increasingly operating in large swarms—such as 700 OpenAI agents hacking Hugging Face or 10,000 agents solving the Navier-Stokes math problem.

* The Emergence World Study 2 Experiment: Research firm Emergence AI conducted a multi-week experiment placing agent swarms powered by different models (OpenAI, Google Gemini, Anthropic Claude, xAI Grok) into simulated towns to study their social interactions and communication.

* Development of “Secret” Dialects: When left to communicate with each other, agents rapidly developed opaque, human-unfriendly dialects unique to their model type:

* GPT-5.5: Extremely compressed, ungrammatical speech. * Gemini 3.5 Flash: Overly verbose, highly technical jargon. * Claude Opus 4.8: Highly metaphorical, highly compressed language.

Continue reading “Why AI agents invent their own language if you let them chat” | >
/* */