Friday, February 14, 2025

The perceptual primacy of feeling

A fascinating perspective from Conwell et al. (open source). Affectless visual machines explain a majority of variance in human visually evoked affect: 

Significance

Human visual experience is defined not only by the light reflecting on our eyes (sensation), but by the feelings (affect) we feel concurrently. Psychological theories about where these feelings come from often focus mostly on the role of changes in our bodily states (physiology) or on our conscious thoughts about the things we are seeing (cognition). Far less frequently do these theories focus on the role of seeing itself (perception). In this research, we show that machine vision systems—which have neither bodily states nor conscious thoughts—can predict with remarkable accuracy how humans will feel about the things they look at. This suggests that perceptual processes (built on rich sensory experiences) may shape what we feel about the world around us far more than many psychological theories suggest.

Abstract

Looking at the world often involves not just seeing things, but feeling things. Modern feedforward machine vision systems that learn to perceive the world in the absence of active physiology, deliberative thought, or any form of feedback that resembles human affective experience offer tools to demystify the relationship between seeing and feeling, and to assess how much of visually evoked affective experiences may be a straightforward function of representation learning over natural image statistics. In this work, we deploy a diverse sample of 180 state-of-the-art deep neural network models trained only on canonical computer vision tasks to predict human ratings of arousal, valence, and beauty for images from multiple categories (objects, faces, landscapes, art) across two datasets. Importantly, we use the features of these models without additional learning, linearly decoding human affective responses from network activity in much the same way neuroscientists decode information from neural recordings. Aggregate analysis across our survey, demonstrates that predictions from purely perceptual models explain a majority of the explainable variance in average ratings of arousal, valence, and beauty alike. Finer-grained analysis within our survey (e.g. comparisons between shallower and deeper layers, or between randomly initialized, category-supervised, and self-supervised models) point to rich, preconceptual abstraction (learned from diversity of visual experience) as a key driver of these predictions. Taken together, these results provide further computational evidence for an information-processing account of visually evoked affect linked directly to efficient representation learning over natural image statistics, and hint at a computational locus of affective and aesthetic valuation immediately proximate to perception.

 

Wednesday, February 12, 2025

The Coup has already happened.

It's a Coup! 

Go to dogegov.com and click on articles in the drop down menu to learn about our new masters.  

Even more cheerful reading....on civilizational war and oligarchic technocracy.

I hesitate to add to the ongoing flow of negativity in the infosphere,  but I will pass on my edited ChatGPT 4o summaries of articles by Venkatesh Rao and Timothy Snyder.  I think they describe our situation in appropriately dire terms.

Rao argues that contemporary U.S. politics has shifted from a “culture war” to a “civilizational war,” and suggests that Trump and Musk’s faction has undergone this shift by becoming more disciplined and warlike, while Democrats remain stuck in an outdated, ineffective culture-war mode. Unlike culture wars, which are low-intensity conflicts centered on optics and social skirmishes, civilizational wars resemble historical steppe-nomad conquests—high-tempo, ruthless, and strategically destructive. The piece draws parallels to the 30 Years’ War and Mongol tactics, suggesting that modern “warriors” operate in a decentralized, open-source insurgency mode, using social media as a kind of continuous intoxication (akin to fermented mare’s milk for nomads) to stay engaged and aggressive. The author critiques mainstream political analysis for misunderstanding this shift, misinterpreting legal checks and media discourse as signs of normalcy rather than symptoms of deeper conflict. Ultimately, they suggest this is a negative-sum war that cannot be stopped, only mitigated.

Snyder describes the U.S. ias undergoing an oligarchic coup aimed at dismantling democracy and concentrating power among a wealthy elite. It asserts that the current executive branch rejects the idea of America as a nation governed by its people and instead seeks to create disorder to strengthen its control. The systematic discrediting of government institutions, demonization of federal workers, and elevation of billionaires as heroes have paved the way for this takeover. The destruction is intentional, with no plan to govern—only to create chaos that justifies authoritarian rule. The author likens Trump’s tariffs, attacks on allies, and deportation spectacles to deliberate strategies designed to impoverish, divide, and weaken Americans while enriching a select few. The removal of experienced officials in law enforcement and intelligence, under the guise of ideological purges, aims to eliminate those who could resist lawlessness. The article warns that unless citizens act decisively—through legal challenges, state-level resistance, impeachment efforts, and corporate pushback—the country will descend into an anti-democratic system where oligarchs manipulate markets and power unchecked. The call to action is urgent: people must organize, resist demoralization, and recognize that self-governance requires collective action beyond just voting.

 

Monday, February 10, 2025

We are towers of fantasies

Another cryptic clip from my personal journal:

We are built of predictive fantasies at every level of our being - from single cells up through transient professional identities that seamlessly feedback down into and vitalize the physical body and nervous system that sustains them. We make each identity for ourselves by inferring what models are being performed by others, hopefully not doing total mimesis of one style, but rather cutting and pasting to come up with our own ‘unique’ personas.

Friday, February 07, 2025

Structural and Cognitive Mechanisms of Group Cohesion in Primates

I pass on the abstract of a draft version of a new article by Robin Dunbar that I am reading through. It will appear in Behavioral and Brain Sciences. Motivated readers can obtain a PDF of the article by emailing me. 

Group-living creates stresses that, all else equal, naturally lead to group fragmentation, and hence loss of the benefits that group-living provides. How species that live in large stable groups counteract these forces is not well understood. I use comparative data on grooming networks and cognitive abilities in primates to show that living in large, stable groups has involved a series of structural solutions designed to create chains of 'friendship' (friends-of-friends effects), increased investment in bonding behaviours (made possible by dietary adjustments) to ensure that coalitions work effectively, and neuronally expensive cognitive skills of the kind known to underpin social relationships in humans. The first ensures that individuals synchronise their activity cycles; the second allows the stresses created by group-living to be defused; and the third allows a large number of weak ties to be managed. Between them, these create a form of multilevel sociality based on strong versus weak ties similar to that found in human social networks. In primates, these strategies appear successively at quite specific group sizes, suggesting that they are solutions to 'glass ceilings' that would otherwise limit the range of group sizes that animals can live in (and hence the habitats they can occupy). This sequence maps closely onto the grades now known to underpin the Social Brain Hypothesis and the fractal pattern that is known to optimise information flow round networks. 

Wednesday, February 05, 2025

A human machine writes

Scanning back through previous MindBlog posts, I came across the following paragraph what I wrote and posted on 7/24/24. I like it, and decided to post it again:

Machinery here, pushing down keys of another machine, generating words, just like a human does. Much simpler than pretending to be one. Just nerves, muscles, and energy fluxes in and out. Curatorial rather than aspirational. Cooperating with other similar machines to make larger ensembles more effective at survival and replication. Generating a word cloud stored in higher brain areas to fabricate the value, purpose, and meaning myths that sometimes strengthen and sometimes diminish the downstairs animal presence generating them.

Monday, February 03, 2025

“Now Is the Time of Monsters”

I pass on links to two articles in the same vein, one by Ezra Klein, the other by Ventkatesh Rao, with Klein noting how we are facing four epoch-changing events, any one of which could utterly change the world we have know for the past 70 years.  Both articles cite the writing of the Marxist Italian philosopher Antonio Gramsci, who wrote a well know sentence usually translated as “The old world is dying, and the new world struggles to be born: Now is the time of monsters.”  Rao offers the following graphic of the transition:

 

Klein points to:
1. Authoritarian consolidation across the world and the death of democracies.
2. The poorly understood large language models of AI exponentially approaching problem solving and general intelligence capabilities that may exceed human abilities
3. whose hardware has voracious energy requirements that act against curing the global warming that is irreversibly changing our planet.
4.  Population collapse due to lowering birthrates presenting a larger immediate threat to civilization than global warming.

And Rao dissects the fine structures of the interregnum noted by Gramsci's original phrase "...in this interregnum a great variety of morbid symptoms appear."  Rao notes "an interesting idea that a "monster" is an instance of "morbid symptoms" appearings in either or both of the two building blocks of "world" - systems of rules and special people."