Lightnews — Scholar-powered news

Splainer (Material Conditions Acknowledger)

@splainer.bsky.social

But a pessimistic-prime phrasing leads Gemini to give an internally inconsistent answer: N_e is "almost always considerably smaller" than N_c, researcher consensus says N_e was likely "substantially larger" than N_c, and yet N_e's not a reliable lower bound.

$Screenshot of the top of the Google search results for the query "Are effective population sizes from genetic studies unreliable as a lower bound on prehistoric human population?", showing the first two paragraphs and inter-paragraph heading of the AI Overview: "Yes, effective population sizes ($\mathbf{N}_{\mathbf{e}}$) derived from genetic studies are generally considered unreliable as a direct lower bound for the actual prehistoric human census population size ($\mathbf{N}_{\mathbf{c}}$). $\mathbf{N}_{\mathbf{e}}$ is almost always considerably smaller than the census size, often by a factor of 10 or more, reflecting the number of individuals contributing to long-term genetic diversity rather than the total number of living individuals. Why Effective Population Size is a Lower Estimate The effective population size is a theoretical value representing the size of an idealized, randomly mating population that would experience the same rate of genetic drift or inbreeding as the real, more complex population. The actual census size is typically much larger due to a variety of biological and demographic factors not accounted for in the idealized model. "$ $Screenshot of the final heading and paragraph of Google's AI Overview for a search with the query "Are effective population sizes from genetic studies unreliable as a lower bound on prehistoric human population?". The heading says "The Key Takeaway" and the paragraph says "This $\mathbf{N}_{\mathbf{e}}$ does not mean the actual human population was ever that small in total. Instead, the consensus among researchers is that the actual census population ($\mathbf{N}_{\mathbf{c}}$) was likely substantially larger, potentially in the range of 100,000 to 300,000 individuals or more. Therefore, the $\mathbf{N}_{\mathbf{e}}$ value serves as an indicator of the amount of genetic diversity maintained by the population, not as a reliable lower bound for the total number of people alive at any given time."$

November 25, 2025 at 5:58 PM

Splainer (Material Conditions Acknowledger)

@splainer.bsky.social

Experimenting again now, I find that the automatically triggered version of Gemini comes to inconsistent conclusions depending on the exact phrasing of the query.

Here one phrasing leads Gemini to give an internally consistent answer: yes, N_e works as a lower bound on N_c.

Screenshot of the top of the results of a Google search. The full search query is shown in a multi-line text box at the top: "Could past effective population sizes estimated from human genomes be used as a lower bound on actual prehistoric human population sizes?", below which is Gemini-generated text ending in the heading "Why It Matters" and the paragraph:

"Using N_e
as a lower bound for prehistoric human populations provides a conservative estimate of the minimum number of individuals that likely existed. While the actual number of people walking the earth was likely higher, the genetic data is often insufficient to resolve many details of Pleistocene human population change. The consistency with which
N_e is smaller than
N_c across various species makes it a reliable minimum value for demographic studies. For instance, a commonly used rule of thumb in conservation biology is that the
N_e/N_c ratio is approximately 0.1 in wild populations.".

November 25, 2025 at 5:53 PM

Splainer (Material Conditions Acknowledger)

@splainer.bsky.social

And yet the Reactionary In Chief is also an affordability-crisis denier like Stancil.

Screenshot of https://x.com/wideofthepost/status/1988590649722216689, @wideofthepost's November 12, 2025 xeet "i've heard this argument before" with an edited image of Will Stancil outside in a brown jacket and his face semi-morphed toward Donald Trump's and quotexeeting @NewsWire_US's headline "There is no affordability crisis, Trump officials insist — Axios".

November 22, 2025 at 12:31 PM

Splainer (Material Conditions Acknowledger)

@splainer.bsky.social

November 19, 2025 at 1:32 AM

Splainer (Material Conditions Acknowledger)

@splainer.bsky.social

It's a fair question.

November 19, 2025 at 12:02 AM

Splainer (Material Conditions Acknowledger)

@splainer.bsky.social

I don't believe you, Misha!!

A poster for Mikhail Gorbachev's 80th birthday gala ("VERTU PRESENTS THE MAN WHO CHANGED THE WORLD"). The headline attendees (apart from Gorbachev himself, natch) are "BRYAN FERRY", "PAUL ANKA", "MILLA JOVOVICH", "KATHERINE JENKINS", "THE SCORPIONS", "MELANIE C", "VALERY GERGIEV", and hosts "SHARON STONE & KEVIN SPACEY".

November 18, 2025 at 11:29 PM

Splainer (Material Conditions Acknowledger)

@splainer.bsky.social

Screenshot of the top of Rudolf Carnap's Wikipedia article, including the redirection warning: "'Carnap' redirects here. For the crime, see Carjacking.".

November 18, 2025 at 1:02 AM

Splainer (Material Conditions Acknowledger)

@splainer.bsky.social

There's a definite tension between what I might call consumer-oriented "leftism" and labor-oriented leftism. It's just that Stancil is incapable of a serious treatment of that tension; he is just too often, too hypocritically, too obnoxiously wrong (and in denial about it).

Screenshot of @ethnyagami's October 2020 tweet "once i touch a lil bit of money this socialist act is over 😂😭😭💯👌🏾".

November 17, 2025 at 11:15 PM

Splainer (Material Conditions Acknowledger)

@splainer.bsky.social

The line chart "Homicide rates have dropped dramatically over the centuries" (from https://ourworldindata.org/data-insights/homicide-rates-have-declined-dramatically-over-the-centuries). It indeed shows homicide rates dropping dramatically, from 7–75 per 100,000 people per year before the 16th century, to less than 4 per 100,000 people per year in the 21st century.

November 17, 2025 at 10:47 PM

Splainer (Material Conditions Acknowledger)

@splainer.bsky.social

As for "serious questions about quality etc.", well, you _can_ cheap out and try Walmart, and the sticker price might be less after you adjust for inflation, but then:

Screenshot of https://www.walmart.com/search?q=1920s-style+shoes, showing the first 3 hits, two of which are unavailable, the third of which is "Women Ballroom Dance Shoes Latin Salsa Closed Toe 1920s Retro Sequins Party Wedding Heels Performance Dancing Pumps", selling for $12.01 plus $5.99 shipping.

Screenshot of the only rating and review (which is of a "Verified Purchase") of the $12.01 shoes in the other image. The rating is 1 star, the review "first of all the shoes are too small and they are peeling and i want my money back please thank you".

November 17, 2025 at 10:00 PM

Splainer (Material Conditions Acknowledger)

@splainer.bsky.social

In fact 1920s-style shoes now cost upward of $200. Halve(-ish) that if they're on sale.

Screenshot of https://www.americanduchess.com/collections/1920s-shoes-and-boots, showing 8 types of "Edwardian Oxfords", "Paris Boots", "Victorian Button Boots", and "Edwardian Leather Shoes", all with a full retail price of $205–$285, two of them on sale for $115.

November 17, 2025 at 9:57 PM

Splainer (Material Conditions Acknowledger)

@splainer.bsky.social

Those shoes in the catalog cost $1.19–$2.98, not $119–$298.

In today's money, $21–$54, not $2100–$5400.

Those shoes were not, in fact, especially expensive!

Screenshot of the Bureau of Labor Statistics CPI Inflation Calculator (https://www.bls.gov/data/inflation_calculator.htm), with $1.19 in November 1925 dollars entered into it, producing a buying-power estimate of $21.49 for September 2025 (most recent month with available data).

November 17, 2025 at 9:23 PM

Splainer (Material Conditions Acknowledger)

@splainer.bsky.social

When you've lost LOLGOP...

Screenshot of two consecutive blueskeets in a thread from an hour ago.

1. Will Stancil (@whstancil.bsky.social): "In a few years we’re going to look back on the “affordability crisis” of 2025 as an embarrassment and everyone will pretend they thought it was ridiculous all along [screenshot of the headline 'A growing number of six-figure earners say the American Dream feels out of reach'"

2. LOLGOP (@thefarce.org): "have you talked to your friends with kids about this certainty of yours?"

November 17, 2025 at 8:18 PM

Splainer (Material Conditions Acknowledger)

@splainer.bsky.social

November 17, 2025 at 8:09 PM

Splainer (Material Conditions Acknowledger)

@splainer.bsky.social

Economic issues, and also defense-of-democracy issues, especially in the late campaign.

Bar chart, with trendlines, from https://images.jacobinmag.com/wp-content/uploads/2024/11/27085907/IMG_8026-1-scaled.jpg. It shows the frequency per thousand words of democracy-related terms (blue bars and blue trendline), and anti-elite terms (brown bars and brown trendline), in the Harris campaign's speeches, rallies, press gaggles, and interviews, week by week from July 17 through November 5.

The democracy-related trendline starts at around 3 mentions per 1000 words in the first week, declines to about 2 mentions per 1000 in September, then rises to about 5 mentions per 1000 in the last campaign week (a higher frequency than the final week's rate of references to the economy, not shown). Chart embedded in https://jacobin.com/2024/11/harris-campaign-economic-populism-democracy.

November 17, 2025 at 12:06 AM

Splainer (Material Conditions Acknowledger)

@splainer.bsky.social

November 16, 2025 at 8:54 PM

Splainer (Material Conditions Acknowledger)

@splainer.bsky.social

Took a few hours but I caught the block for that, so for posterity: Andrew Parker @andrewparker.bsky.social) was the crank misrepresenting @garyalexander.bsky.social.

#YouCanBlockButYaCantHide

Screenshot of @andrewparker.bsky.social's quoteblueskeet "A good object lesson in how easy it is even for self-described leftists to fall into the conservative mindset of “nonono, there’s no systemic discrimination, individual quality is what decided those outcomes”", quoting @garyalexander.bsky.social's blueskeet "Personally, I think the idea that all women should be measured by the failures of Hillary Clinton and Kamala Harris is actually a sort of sexist doomerism.".

November 16, 2025 at 8:43 PM

Splainer (Material Conditions Acknowledger)

@splainer.bsky.social

Latest Harvard CAPS / Harris(X) poll shows self-described liberals split approximately evenly (54/46) on open borders, which is a less extreme position than the no-borders position that's allegedly "literally the liberal position".