Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog

Category: mapping

Epistemological Fault Lines Between Human and Artificial Intelligence

Walter (Dated: December 22, 2025)

See… https://osf.io/preprints/psyarxiv/c5gh8_v1

Abstract: Large language models (LLMs) are widely described as artificial intelligence, yet their epistemic profile diverges sharply from human cognition. Here we show that the apparent alignment between human and machine outputs conceals a deeper structural mismatch in how judgments are produced. Tracing the historical shift from symbolic AI and information filtering systems to large-scale generative transformers, we argue that LLMs are not epistemic agents but stochastic pattern-completion systems, formally describable as walks on high-dimensional graphs of linguistic transitions rather than as systems that form beliefs or models of the world. By systematically mapping human and artificial epistemic pipelines, we identify seven epistemic fault lines, divergences in grounding, parsing, experience, motivation, causal reasoning, metacognition, and value. We call the resulting condition Epistemia: a structural situation in which linguistic plausibility substitutes for epistemic evaluation, producing the feeling of knowing without the labor of judgment. We conclude by outlining consequences for evaluation, governance, and epistemic literacy in societies increasingly organizedaround generative.

Cc: ronald cicurel ernest davis amitā kapoor darius burschka william hsu moshe vardi luis lamb jelel ezzine amit sheth bernard W. kobes.

See…

New dataset maps global city boundaries in high resolution from 2000 to 2022

A research team led by Prof. Liu Liangyun from the Aerospace Information Research Institute of the Chinese Academy of Sciences (AIRCAS) has produced the first comprehensive, high-resolution map of global city and town boundaries, offering a view of how urban boundaries have expanded and transformed over the past two decades. The new dataset—derived from 30-meter-resolution satellite observations—fills a long-standing gap in global urban studies.

Cities and towns are the dominant form of human settlement, playing a crucial role in sustaining ecological balance and advancing sustainable development. However, their complex spatial structures and rapid evolution have made high-resolution global urban boundary datasets scarce. To address this gap, the team integrated the GISD30 global impervious surface dynamic dataset with LandScan global population data to develop the Global City and Town Boundaries (GCTB) Dataset, which covers the period from 2000 to 2022.

Published in Scientific Data, the study details the researchers’ development of a morphology-oriented boundary delineation framework that combines kernel density estimation (KDE) and cellular automata (CA) to accurately map urban boundaries. When compared with multiple reference datasets, the GCTB Dataset showed the strongest agreement with the manually curated Atlas of Urban Expansion, achieving an R2 value of approximately 0.88—indicating high reliability in capturing urban extents.

TransBrain: a computational framework for translating brain-wide phenotypes between humans and mice

TransBrain translates brain phenotypes between mouse and human via homology mapping, thus making it possible to capitalize on the wealth of knowledge about the mouse brain and gain insights into the human brain.

NASA’s Roman telescope will observe thousands of newfound cosmic voids

Our universe is filled with galaxies, in all directions as far as our instruments can see. Some researchers estimate that there are as many as 2 trillion galaxies in the observable universe. At first glance, these galaxies might appear to be randomly scattered across space, but they’re not. Careful mapping has shown that they are distributed across the surfaces of giant cosmic “bubbles” up to several hundred million light-years across. Inside these bubbles, few galaxies are found, so those regions are called cosmic voids. NASA’s Nancy Grace Roman Space Telescope will allow us to measure these voids with new precision, which can tell us about the history of the universe’s expansion.

“Roman’s ability to observe wide areas of the sky to great depths, spotting an abundance of faint and distant galaxies, will revolutionize the study of cosmic voids,” said Giovanni Verza of the Flatiron Institute and New York University, lead author on a paper published in The Astrophysical Journal.

Cosmic recipe The cosmos is made of three key components: normal matter, dark matter, and dark energy. The gravity of normal and dark matter tries to slow the expansion of the universe, while dark energy opposes gravity to speed up the universe’s expansion. The nature of both dark matter and dark energy is currently unknown. Scientists are trying to understand them by studying their effects on things we can observe, such as the distribution of galaxies across space.

Rethinking the centrality of brain areas in understanding functional organization

For decades, neuroscience textbooks have taught us that the brain is organized into discrete areas — like Broca’s area for language or V1 for early vision, each with a well-defined role. This kind of areal parcellation has shaped how we interpret brain imaging, neural recordings, and even theories of cognition.

But this new article challenges that foundational idea. Instead of treating brain areas as the central units of brain function, the authors argue that brain organization is more complex, multi-layered, and distributed than traditional area-based frameworks suggest.

The authors begin with a simple observation: the ways in which neuroscientists define cortical areas, based on cell structure (cytoarchitecture), connectivity, or response properties — don’t always point to the same boundaries. In other words, different methods of dividing the cortex produce different “maps,” and there’s surprisingly little convergence on a single, definitive set of brain areas.

This inconsistency raises a big question: If areas aren’t consistently defined by structure or connectivity, can we really treat them as the fundamental units of brain function.

Parcellation of the cortex into functionally modular brain areas is foundational to neuroscience. Here, Hayden, Heilbronner and Yoo question the central status of brain areas in neuroscience from the perspectives of neuroanatomy and electrophysiology and propose an alternative approach.

Study reveals visual processing differences in dyslexia extend beyond reading

New research published in Neuropsychologia provides evidence that adults with dyslexia process visual information differently than typical readers, even when viewing non-text objects. The findings suggest that the neural mechanisms responsible for distinguishing between specific items, such as individual faces or houses, are less active in the dyslexic brain. This implies that dyslexia may involve broader visual processing differences beyond the well-known difficulties with connecting sounds to language.

Dyslexia is a developmental condition characterized by significant challenges in learning to read and spell. These difficulties persist despite adequate intelligence, sensory abilities, and educational opportunities. The most prominent theory regarding the cause of dyslexia focuses on a phonological deficit. This theory posits that the primary struggle lies in processing the sounds of spoken language.

According to this view, the brain struggles to break words down into their component sounds. This makes mapping those sounds to written letters an arduous task. However, reading is also an intensely visual activity. The reader must rapidly identify complex, fine-grained visual patterns to distinguish one letter from another.

3D maps reveal hidden microenvironments shaping mouse brain connectivity

Recent technological and scientific advances have opened new possibilities for neuroscience research, which is in turn leading to interesting new discoveries. Over the past few years, many groups of neuroscientists worldwide have been trying to map the structure of the brain and its underlying regions with increasing precision, while also probing their involvement in specific mental functions.

As mapping the human brain in detail is often challenging and requires significant resources, many studies focus on other mammals, particularly mice or other rodents. Most mouse brain atlases delineated to date map the density of neurons or other brain cells (i.e., how many cells are packed in specific parts of the brain). In contrast, fewer works also tried to map the shape of neurons in the mouse brain and interactions between them.

Researchers at Fudan University and Southeast University recently set out to map dendrites (i.e., branch-like extensions of neurons via which they receive signals from other cells) in the mouse brain. Their paper, published in Nature Neuroscience, unveils groups of structures in the mouse brain that influence how neurons function and connect to other neurons, also known as microenvironments.

Astrocyte diversity across space and time charted in new atlas

When it comes to brain function, neurons get a lot of the glory. But healthy brains depend on the cooperation of many kinds of cells. The most abundant of the brain’s non-neuronal cells are astrocytes, star-shaped cells with a lot of responsibilities. Astrocytes help shape neural circuits, participate in information processing, and provide nutrient and metabolic support to neurons. Individual cells can take on new roles throughout their lifetimes, and at any given time, the astrocytes in one part of the brain will look and behave differently than the astrocytes somewhere else.

After an extensive analysis by researchers at MIT, neuroscientists now have an atlas detailing astrocytes’ dynamic diversity. Its maps depict the regional specialization of astrocytes across the brains of both mice and marmosets—two powerful models for neuroscience research—and show how their populations shift as brains develop, mature, and age.

The open-access study, reported in the Nov. 20 issue of the journal Neuron, was led by Guoping Feng, the James W. (1963) and Patricia T. Poitras Professor of Brain and Cognitive Sciences at MIT.

How Agentic BAS AI Turns Threat Headlines Into Defense Strategies

Picus Security explains why relying on LLM-generated attack scripts is risky and how an agentic approach maps real threat intel to safe, validated TTPs. Their breakdown shows how teams can turn headline threats into reliable defense checks without unsafe automation.

Comprehensive map reveals neuronal dendrites in the mouse brain in greater detail

Understanding the shape or morphology of neurons and mapping the tree-like branches via which they receive signals from other cells (i.e., dendrites) is a long-standing objective of neuroscience research. Ultimately, this can help to shed light on how information flows through the brain and pin-point differences associated with specific neurological or psychiatric disorders.

The X. William Yang Lab at the Jane and Terry Semel Institute and the Department of Psychiatry and Biobehavioral Sciences at University of California, Los Angeles (UCLA) have devised new sophisticated methods to map neuronal dendrites in the mouse brain, which combine large-scale data collection with genetic labeling techniques and computational tools.

Their research approach, outlined in a paper published in Nature Neuroscience, allowed them to create a comprehensive map of two genetic types of neurons in the mouse brain, known as D1-and D2-type striatal medium spiny neurons (MSNs).

/* */