Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog

Category: neuroscience – Page 87

AI-Discovered Cognitive Models Reveal Novel Insights into Human and Animal Learning

The problem? Human brains (and animal brains, too) are incredibly complex. While these handcrafted models are great starting points, they often oversimplify things and miss the messy, rich reality of actual behavior. On the flip side, using powerful, flexible AI to analyze data can capture that richness, but AI usually gives us a “black box”—it finds patterns but can’t explain *why* or *how* it found them, leaving scientists to do the heavy lifting of figuring out the rules.

Scientific models are widely used across the natural sciences as an interface between scientific theories and empirical data [1]. Such models play a key role, for example, in the study of human and animal learning, where they express algorithmic hypotheses and relate them to psychology and neuroscience data [2, 3]. These models are traditionally handcrafted by expert researchers based on existing theory or new insights. Such handcrafted models, however, are now known to fall short of capturing the full richness of behavior, even in their narrow domains [4– 7]. An alternative data-driven approach has emerged, seeking to discover new insights by fitting and interpreting flexible models [8– 11]. However, these tools require substantial human effort to derive insight from data, and it has been unclear how to discover new ideas from data efficiently. Here, we present DataDIVER, a general approach for automatically discovering computational models from data, and demonstrate that these models surface novel mechanistic insights into human and animal learning. Our approach delivers models that take the form of short computer programs, which are optimized both to fit data well and to be simple. These programs explicitly connect with existing theoretical frameworks and are readily understandable by human scientists. They can also be used to make novel predictions, some of which we show are borne out in re-analysis of existing data. General-purpose tools for surfacing new ideas from data, especially in combination with the large datasets that are increasingly available in many fields, stand to dramatically accelerate scientific discovery.

The authors have declared no competing interest.

Precognition, Retrocausation, and the Unconscious with Eric Wargo

Eric Wargo, PhD, is author of Time Loops: Precognition, Retrocognition, and the Unconscious. He is an anthropologist and science writer. His blog is http://thenightshirt.com/.

Here he defines time loops as akin to self-fulfilling prophecies. He asserts that they could be the very basis of the creative process. He explains that retrocausation is to physics what precognition is to parapsychology. He explores the social and psychological dynamics associated with the notion of premonitions. He reviews the experiment in time of J. W. Dunne suggesting that dreams contain much information seemingly derived from the future. He applies Dunne’s methology to dreams of Sigmund Freud.

New Thinking Allowed host, Jeffrey Mishlove, PhD, is author of The Roots of Consciousness, Psi Development Systems, and The PK Man. Between 1986 and 2002 he hosted and co-produced the original Thinking Allowed public television series. He is the recipient of the only doctoral diploma in \.

Critical Thalamocortical Coordination Dynamics Track Conscious State Transitions

Abstract Despite substantial progress in identifying neural correlates of consciousness, no unified quantitative framework currently derives a formally specified order parameter for conscious-state organisation from established neurophysiological principles, or links thalamocortical coordination dynamics to measurable state transitions across pharmacological, pathological, and perturbational conditions through a single computational formalism. We propose a neurocomputational theoretical framework in which conscious states are associated with metastable regimes of large-scale thalamocortical coordination operating near critical dynamical boundaries. The framework is formalised through a dynamic coordination functional Φ(t), defined as a surface integral over the thalamocortical interface and directly operationalisable from high-density EEG as a weighted combination of gamma-band power spectral density, thalamocortical coherence, and theta-gamma phase-amplitude coupling. The thalamic reticular nucleus (TRN) is identified as the anatomical implementation of the control parameter governing proximity to the critical point, grounded in a Wilson-Cowan model of TRN inhibitory gating whose bifurcation structure is characterised computationally. Numerical simulation of the linearised field equation on the thalamocortical boundary demonstrates internal consistency: the simulated system produces power-law recovery dynamics tau_rec proportional to | θ — θ _c|^v with nu consistent with model A universality class [0.5, 1.5], and a Kuramoto mean-field derivation establishes that Φ(t) emerges as the natural order parameter of coupled thalamocortical oscillators rather than being postulated. The joint (|Φ(t)|, Var[|Φ(t)|]) phase space correctly separates simulated waking, anaesthetic, ictal, and minimally conscious regimes without parameter fitting to empirical data. All simulation code is publicly available. Six quantitatively specific, independently falsifiable predictions are derived across five experimental domains: power-law Gamma Dip scaling in near-threshold EEG with a specific exponent range; causal disruption of thalamocortical coherence by selective TRN silencing; opposite EEG scaling exponent deviations in ASD versus schizophrenia; systematic Φ_est collapse under propofol anaesthesia correlated with PCI; Φ_est as a real-time consciousness biomarker in disorders of consciousness; and clinical validity of Φ_est in disorders of consciousness and ictal state discrimination by the metastability index. Each prediction is stated with quantitative thresholds and a pre-specified falsification criterion. The framework provides: the first anatomically specified and formally derived order parameter for conscious-state organisation directly operationalisable from passive EEG; a mechanistically grounded identification of the TRN as the dynamical control parameter, testable by a single optogenetic experiment; and a computationally validated, pre-registerable programme of six falsifiable predictions defining a tractable empirical agenda. Φ_est would constitute a candidate real-time consciousness biomarker if the framework’s predictions are confirmed in purpose-designed experiments.

NIH-funded study suggests that testosterone suppresses brain tumor growth in males

Findings may warrant exploration of the hormones as glioblastoma treatment.

In a new National Institutes of Health (NIH)-funded study, scientists at Cleveland Clinic discovered that hormones associated with male development may play a key role in limiting the growth of brain tumors in men. The research team found that the loss of androgen hormones, such as testosterone, in a preclinical model of glioblastoma drove tumor growth by inducing local inflammation and triggering the production of stress hormones. In an analysis of data from more than 1,300 men with glioblastoma, the authors found that supplemental testosterone was significantly associated with improved survival, which was consistent with their preclinical experiments.

“This outcome is a welcome surprise and may potentially offer a lead for new treatments for a kind of cancer that is deadlier in men,” said Anthony Letai, M.D., Ph.D., director of NIH’s National Cancer Institute (NCI).

Britain’s oldest cave art may have been rediscovered in Bacon Hole cave

The oldest cave art in Britain may have been discovered, or more likely rediscovered, in a cave on the Gower Peninsula in South Wales, possibly dating back around 17,000 years.

The red stripe markings on the walls of a cave called Bacon Hole were first spotted in 1912. They were hailed by their finders, Professor William Sollas and Henri Breuil, as the first Paleolithic cave art in the British Isles. However, skeptics dismissed the discovery, arguing that the markings were merely natural mineral deposits.

Over time, the claim and subsequent debate drifted out of public consciousness and have largely been forgotten about. That was until scientists decided to reexamine the marks.

New study suggests the brain applies different standards of beauty to paintings and architecture

A recent study published in the journal Psychology of Aesthetics, Creativity, and the Arts provides evidence that the human brain applies different standards of beauty depending on the type of visual art it evaluates. By comparing the visual properties of historical building facades and artistic paintings, scientists found that architects and painters weigh aesthetic features like symmetry and complexity quite differently.

When people look at an image, their appreciation of its beauty relies on several visual variables. These variables include properties such as color, balance, symmetry, complexity, and the relationship between the main subject and its background. Psychological theories of visual perception propose that humans tend to prefer sensory properties that the brain can process easily.

“I have been interested in the Valuation System of the brain, the network that learns and deploys values for decision-making,” said Norberto Grzywacz, a professor of psychology at Loyola University Chicago. “In particular, I have had interests in aesthetic values, which this system also processes. At some point, I asked myself whether aesthetic values in a sensory domain, for example vision, are universal or specific to different domains.”