Last installment of the network science series. Forget the AI takeover, we're already mindless automatons (but don't tell the free will enthusiasts).
A physics explanation shows why US elections keep ending 50:50—and why more spending won't change that
Apr 2026, phys.org
A spending threshold in US House races of roughly 1.8 million USD per campaign limits outcomes. Below it, social dynamics shape outcomes. Above it — on both sides — elections systematically trend toward a draw, no matter how much either party ultimately spends, while driving polarization higher. ... When both parties spend over 1.8 million USD, social influence becomes negligible and the election very often ends in a close race.Further, on incumbents: The researchers put a number on this structural advantage. Even if the incumbent spends nothing, a challenger must invest roughly 140,000 USD just to neutralize the baseline incumbency effect. When the incumbent spends around 900,000 USD, the challenger still faces a disadvantage equivalent to about 20% of total campaign cost, purely as a consequence of the system's phase structure, not the incumbent's individual qualities.
via Complexity Science Hub Vienna: Jan Korbel et al, Empirical Validation of the Polarization Transition in a Double-Random Field Model of Elections, Physical Review Letters (2026). DOI: 10.1103/9gjj-1df6.
On arXiv: DOI: 10.48550/arxiv.2510.00612
Image credit: A fungus Talaromyces purpureogenus known for its red, diffused pigment
Wim van Egmond - Nikon Small World Photomicrography Competition - 2025
From public kissing to talking during movies, a simple formula predicts moral norms across cultures
Apr 2026, phys.org
"An implication of our simple formula is that norms for one behavior can inform us about norms for a very different behavior. For example, the more okay it is to kiss in the street (a behavior that elicits concerns about purity), expect it to be less okay to beat children (which instead elicits concerns about harm)."
Moral Flavors Model = TC(B) + MF(B) x MT(S)
- TC - total concern that behavior B is seen to elicit
- MF - moral flavor either individualizing type (harm, fairness) or binding type (purity, authority, loyalty)
- MT - moral taste measures emphasis of individualizing concerns over binding concerns
via Institute for Future Studies in Sweden: Kimmo Eriksson et al, Same flavours, different taste buds: a theory for predicting social norms for specific behaviours across cultures, Journal of the Royal Society Interface (2026). DOI: 10.1098/rsif.2025.1122.
Post Script - List of Morally Contentious Behaviors
- claiming government benefits to which you are not entitled
- avoiding a fare on public transport
- stealing property
- cheating on taxes
- accepting a bribe
- homosexuality
- prostitution
- abortion
- divorce
- sex before marriage
- suicide
- euthanasia
- for a man to beat his wife
- parents beating children
- violence against other people
- terrorism as a political, ideological or religious mean
- having casual sex
- political violence
- the death penalty
--World Values Survey Wave 7 data (2017–2021) for 42 societies; Minkov M, Kaasa A. 2022 Do dimensions of culture exist objectively? A validation of the revised Minkov-Hofstede model of culture with World Values Survey items and scores for 102 countries. J. Int. Manag. 28, 100971. doi:10.1016/j.intman.2022.100971
How deceptive content reached millions of voters during the 2020 US elections
Apr 2026, phys.org
(Note that Facebook was directly involved in this research, so assume you are being intentionally deceived by this data, at least to some extent, and in an attempt to make Facebook look better than they are)
They focused on 49 deceptive networks that targeted adult Facebook and Instagram users in the US during the 2020 election, both disincentivized networks of users who engaged in inaccurate political discourse and financially motivated networks disseminating content that is largely dismissed as spam or clickbait. 13 out of the 49 identified were "coordinated inauthentic behavior networks", and the remaining 36 networks were found to be financially motivated (by advertising). They were organized by characteristics like where they originated, how many accounts they ran, and what they posted about, as well as by activity and reach.The networks were measured to have reached about 40 million users, or 15% of the overall network, and were highly concentrated - only 3 of the 49 networks accounted for over 70% of all the users reached. One of which was an account called "Rally Forge' created in the US. (It's really fucking frustrating, in this case for example, to try and get the list of the ** other 2 ** networks, but we can't because it's behind a paywall; a paywall that we already paid for with our tax dollars. And yet we then turn around and give it all away for free to the same companies so they can gobble it up into their too fat, too slow, and too stupid artificial intelligence engines.)
So anyway, here's the important part:
Networks reached most of their audience not directly, but because ordinary users — people unaffiliated with the networks — reshared their content. The network with the highest reach, for example, reached about 1.3 million users directly, but 13 million indirectly through reshares by ordinary users. (That's 10 times more people, for the mathematically challenged among us) ... They suggest that interventions that only target deceptive networks might be insufficient, as regular users are also contributing to the dissemination of misleading content.
So, if you are one of these impact layer people who get hit first, and none of us could really know if that's us because the inauthentic group networks are hidden by design, then by simply using the platform, ie sharing articles with your friends, you are doing up to ten times the work of the company, the group, trying to advertise or influence - we are literally working for them, for free, by taking the attention of our friends and giving it to them, so we are exploiting our own social network for their benefit, likely lessening our own social capital for their increasing financial capital
One last thing:
Interestingly, the researchers observed that financially motivated networks, which some previous studies dismissed or considered less impactful in the context of elections, produced a substantial amount of political content. Moreover, the content they disseminated often reached far more users than the posts shared across politically motivated networks. (In other words, election financing things like Citizens United, where anyone, anywhere, using otherwise hidden money, also called dark money, can purchase otherwise democratic election campaigns and the candidates they support.)
via Stanford University, Meta, University of Pennsylvania: Ruth E. Appel et al, How deceptive online networks reached millions in the US 2020 elections, Nature Human Behaviour (2026). DOI: 10.1038/s41562-026-02435-2.
No comments:
Post a Comment