Friday, June 27, 2025

AI is good, actually

 


You preach it, boy!  

Guilty pleasure

 




"Professor" Dave attacks an unhinged Incel and conspiracy theorist who claims to have learned physics "intuitively". I admit watching this was a guilty pleasure. If Mr Farina only attacked people like this, I might almost start to like him. Almost.  

Feralization

 


If you know, you know. Muh-hah-hah-hah! 

How long does it take for a domesticated species of plant to turn wild again? Does this have any bearing on the question of when agriculture began?

Excellent and deep question — let’s tackle it in two parts:


🌱 1. How long does it take for a domesticated plant species to “turn wild” again?

This process is called feralization — when a domesticated species reverts to a wild or semi-wild state after escaping cultivation or human management.

⏳ How fast can it happen?

  • In some cases, feralization can begin within a few generations (e.g., 5–50 years), especially in annual plants that reproduce quickly.

  • Traits that are lost or reduced during domestication — like seed shattering, seed dormancy, or thorn production — can re-evolve under natural selection if beneficial in the wild.

  • Full reversion to a wild-type genome is rare, but feral plants often show:

    • Increased seed dispersal

    • Decreased yield

    • Resumption of natural defenses (e.g., toxins or spines)

📌 Example: Weedy rice (Oryza sativa f. spontanea), which evolved from domesticated rice and re-developed wild traits in only a few centuries (possibly faster in some regions).


📜 2. Does this affect how we date the beginning of agriculture?

Yes — feralization complicates the archaeological record of agriculture in important ways:

🧭 Here's how:

  • Early farming was likely intermittent, small-scale, and low-intensity.

  • If agriculture was abandoned for even a few decades or centuries, domesticated crops could go feral and vanish from the landscape and archaeological visibility.

  • This means that early domestication events could have occurred thousands of years before the earliest definitive archaeological evidence.

🔄 Agriculture as a long process:

  • Modern archaeology sees domestication as a gradual, multi-generational process, not a single moment.

  • Some species show pre-domestication cultivation — early humans planted them before the genetic signs of full domestication appear.

  • If early farming societies collapsed or moved, their plants may have reverted — making it seem like agriculture started “later” than it actually did.

🧬 Genetic evidence shows that domestication traits (like non-shattering ears in wheat or barley) accumulated slowly, and can be lost under feral conditions.


🧾 Summary

✅ Domesticated plants can revert toward wild forms surprisingly quickly (within decades to centuries), depending on the species and environment.

🧠 This means that early, failed, or abandoned attempts at agriculture may have left no clear trace in the archaeological record.

📆 As a result, the true origins of agriculture may be older and more complex than previously thought, with cycles of cultivation, abandonment, and re-domestication.


Would you like examples of specific plants (like wheat, barley, or maize) and how quickly they can feralize or re-domesticate?

Yes, all three. Also rice!

Great — let’s explore how wheat, barley, maize, and rice behave when feralized (i.e., revert from domestication) or re-domesticated, and what that tells us about the timing and complexity of agriculture’s origins.


🌾 1. Wheat (Triticum spp.)

🧬 Domesticated Traits:

  • Non-shattering spikelets (seeds stay on the stalk, making harvesting easier)

  • Larger grains

  • Softer seed coats

🔄 Feralization/Reversion:

  • Under wild conditions, non-shattering becomes a disadvantage (seeds don’t disperse).

  • Studies show that feral wheat lines can evolve partial shattering again within ~20–50 generations (500–1,000 years).

  • However, wheat is less weedy and does not easily naturalize without human help.

🧠 Implication:

Wheat domestication was slow, taking 1,000–2,000+ years, and could reverse partially if early agriculture was abandoned.


🌾 2. Barley (Hordeum vulgare)

🧬 Domesticated Traits:

  • Like wheat: non-shattering spikes, larger grains

  • Shorter dormancy, more uniform germination

🔄 Feralization:

  • Weedy barley populations (H. vulgare f. spontaneum) are common near fields.

  • Barley can hybridize with wild relatives, rapidly reintroducing wild traits like:

    • Seed dormancy

    • Rough awns (barbs)

    • Shattering

📌 Some barley reverts to near-wild form within <500 years when left uncultivated.

🧠 Implication:

Barley is more plastic than wheat, showing frequent gene flow with wild forms. This makes archaeological dating of “fully domesticated” barley complicated.


🌽 3. Maize (Zea mays)

🧬 Domesticated Traits:

  • Derived from teosinte (wild grass from Mexico)

  • Maize has:

    • Larger cobs

    • Multiple rows of kernels

    • Non-shattering seeds (fully dependent on humans to spread)

🔄 Feralization:

  • Maize is highly domesticated and does not easily revert.

  • However, teosinte-maize hybrids can arise spontaneously if maize is grown near wild relatives.

  • Some “wild maize” types in Mexico are likely re-wilded hybrids, but they don’t fully return to teosinte.

🧠 Implication:

Maize is an outlier — it’s so altered that it can’t survive as-is in the wild. But gene flow with teosinte can blur the archaeological record.


🌾 4. Rice (Oryza sativa)

🧬 Domesticated Traits:

  • Shorter seed dormancy

  • Non-shattering grains

  • Compact panicles (grain heads)

  • Reduced awns

🔄 Feralization:

  • Rice rapidly feralizes — a major modern problem known as weedy rice or red rice.

  • This happens through:

    • Escape from fields

    • Hybridization with wild rice (O. rufipogon)

  • Traits like seed shattering, dormancy, and red pericarps reappear in <10 generations.

📌 Weedy rice is genetically part-domesticated but behaves wild — showing how quickly selection pressure can undo human cultivation traits.

🧠 Implication:

Rice demonstrates how fast domesticated traits can be lost in just decades. This has serious implications for interpreting early rice domestication in East and Southeast Asia.


🧾 Overall Takeaways

PlantSpeed of FeralizationKey Points
WheatSlow–moderate (centuries)Stable once fully domesticated; partial reversions
BarleyFast–moderate (decades–centuries)Hybridizes easily; common weedy forms
MaizeRare, via hybridization onlyCannot survive without human aid, but gene flow with teosinte confuses the picture
RiceFast (within decades)Weedy forms arise quickly; feral rice is a major modern issue

🌱 Final Thought

These examples show that:

  • Domestication is reversible, especially in early stages.

  • Many early agricultural episodes could disappear from the archaeological record within centuries, due to feralization.

  • Therefore, true beginnings of agriculture may predate the first archaeobotanical evidence by millennia.

Would you like a diagram or timeline showing how these crops evolved across time and geography? 

Three men in a boat

 

So I re-watched “Jaws”. I never grokked this production the first couple of times I saw it. It hardly scared me and I found it boring! Today, I rather consider it very, very strange. Indeed, it comes across as two entirely different films. The first half is a rather (stereo)typical horror flick with all the usual ingredients. The monster attacks a wholesome all-American resort, munching on stoned hippies, children and stupid hillbillies. There is an idiot mayor, a scientific genius and a stable police officer just trying to do his job. OK, maybe it´s a *bit* original. Like the tiger shark who turns out to be a rare guest from Louisiana!

The second half is the real “Jaws”. The three main characters Brody, Hooper and Quint confront the super-sized shark in a boat that´s obviously too small and fragile (rather than calling the coast guard or navy). Quint turns out to be half-mad and personally obsessed with killing great white sharks. He is the “Captain Ahab” of the story. The whole thing makes zero sense, except as a bizarre male rite of passage. The shark is too large and too intelligent to be a normal animal. Indeed, it seems to be intrinsically *evil*.

Speculations about allegorical meaning are difficult to fend off. The three men on the worthless barge represent different kinds of Americans, perhaps different generations: the old and crazy war veteran Quint (who is presumably working class), the middle-aged and middle-class police officer, and the young well-educated scientist (implied to be upper class). Is this a vision of an America united against its external enemies? (“Jaws” was released in 1975.) Or is the shark a symbol of Nature showing its fangs? 

It´s intriguing to note that the irrational sailor Quint and the “rational” scientist Hooper turn out to be equally crazy (at least after a fashion). It´s also interesting that the person who eventually kills the shark is Brody, the stable White middle-class guy with an official police badge. He does so in the old fashioned way – with a rifle and some explosives – while Quint and Hooper tried various alternative techniques which completely failed. Order has been restored on Amity Island (note the 4th July parade – Amity is of course a symbol for America and its proverbial way of life). At least it´s been temporarily restored until the sequels, but Steven Spielberg had nothing to do with those.

So I suppose “Jaws” is at least somewhat interesting…


An uncomfortable truth

 


Tonight´s AI essay. 

Are there any figures indicating how many workers (or working class voters) voted for Donald Trump in the 2016, 2020 and 2024 US presidential elections?

Yes, although exact numbers can be hard to pinpoint, there are several reliable estimates and studies indicating how many working-class or blue-collar voters supported Donald Trump in the 2016, 2020, and 2024 U.S. presidential elections. The data comes from exit polls, academic studies, and voter file analyses, with "working class" often defined as voters without a four-year college degree or by occupation/income.


🔹 2016 Election

Definition Used: Voters without a college degree (often used as a proxy for working class)

  • Trump won 67% of white non-college voters, according to CNN exit polls.

  • He won about 52% of all voters without a college degree.

  • Among union households, Trump performed surprisingly well: 43% of union households voted for Trump, compared to 51% for Clinton.

  • In Rust Belt states (e.g., Pennsylvania, Michigan, Ohio), white working-class voters were critical to Trump's narrow victories.

Estimate: About 40–45% of the total working-class vote nationwide went to Trump, with white working-class support reaching nearly 2/3.


🔹 2020 Election

  • Trump maintained strong support among white working-class voters, winning 64% of them (Pew Research, validated voter study).

  • His performance among non-white working-class voters improved slightly from 2016, especially among Latino men.

  • Among all voters without a four-year degree, Trump won about 54% (vs. Biden’s 44%).

Estimate: Trump received a slight majority (54%) of working-class votes, especially from white and rural voters. Biden did better with union households than Clinton had.


🔹 2024 Election (Early Data/Estimates)

Because full data is still being collected and analyzed, most numbers come from early exit polls and academic projections.

  • Trump again won a majority of working-class voters (especially non-college whites).

  • According to early Edison exit polls:

    • Trump won 65–70% of white non-college voters.

    • He made further inroads with working-class Latinos and some Black voters, especially men.

  • In states like Michigan and Pennsylvania, working-class turnout was again a key battleground.

Preliminary Estimate: Trump won 55–58% of working-class voters in 2024, showing continued dominance with white working-class voters and modest gains with others.


🧠 Notes on Definitions

  • "Working class" is not a perfect category; some studies use:

    • Educational attainment (no college degree),

    • Occupation (manual labor, service, manufacturing),

    • Income levels (lower-middle to middle income),

    • Or union membership (for organized labor subset).

  • Trump’s rhetoric on trade, immigration, and “America First” themes consistently appealed to parts of the working class who felt economically left behind or culturally alienated.


📊 Summary Table

Election YearWorking-Class Voter Support for TrumpKey Details
2016~45% overall; ~67% white non-collegeHigh swing in Rust Belt; anti-globalization message resonated
2020~54% overall; ~64% white non-collegeStrong retention; slight gains among working-class Latinos
2024~55–58% overall (est.)Strong white working-class base; modest minority gains

If you'd like, I can provide state-by-state breakdowns or focus on union households or non-white working-class voters as subgroups.

There were claims in 2020 (based on opinion polls before the elections) that Trump´s support among White workers were going down. But it seems that he gained more such voters in the elections?

You're absolutely right to point out the discrepancy. In the run-up to the 2020 election, many pre-election opinion polls suggested that Donald Trump was losing support among white working-class voters, particularly in key swing states. However, post-election data from validated voter studies and exit polls tell a different story: Trump actually held onto — and in some cases modestly increased — his support among white working-class voters compared to 2016.

Here’s what happened in more detail:


🔄 Expectations vs Reality: White Working-Class Vote in 2020

🔹 What the Polls Predicted (Pre-Election 2020):

  • Several major polls (e.g. Quinnipiac, NYT/Siena, Pew) suggested erosion in Trump’s support among:

    • White non-college voters, especially women.

    • Union households in Rust Belt states.

  • The narrative was that Biden was doing better with working-class whites than Clinton had.

  • Pollsters cited Trump's perceived mismanagement of COVID, economic uncertainty, and fatigue among moderate voters.

➤ For example, a September 2020 Monmouth poll had Trump trailing among white working-class voters in Wisconsin, which turned out to be incorrect.


🔹 What Actually Happened (Post-Election 2020):

According to Pew Research Center's validated voter study and Edison exit polls:

  • White working-class voters backed Trump by ~64%, very close to his 2016 margin.

  • Trump’s support held steady or even improved in rural areas, and he deepened margins in parts of Ohio, Pennsylvania, and Iowa.

  • Biden improved slightly among white working-class women, but not nearly enough to offset Trump’s dominance with men in that group.

🔍 Pew’s 2020 report: Among white voters without a four-year degree:

  • Trump: 64%

  • Biden: 35%

This was about the same as 2016 (when Trump got 67% and Clinton 28% among that group per CNN exit polls).


🧠 Why Were the Polls Wrong?

Several reasons explain the mismatch:

  1. Polling error among non-college whites:

    • Many pollsters underrepresented or misweighted this group.

    • Some white working-class voters are less likely to respond to polls, skewing samples.

  2. Social desirability bias:

    • Some respondents may not have admitted voting for Trump in polls, especially after media criticism of his presidency.

  3. Turnout models were off:

    • Many polls underestimated the high turnout among rural and working-class whites, especially in red counties.

  4. Late shifts:

    • Trump’s strong messaging on reopening the economy and culture war themes galvanized working-class support in the final weeks.


🧾 Summary

FactorPolls (Pre-2020)Actual Result
White working-class supportPredicted to drop (esp. women)Stayed stable or rose slightly
Swing state working classBiden projected to gain groundTrump retained or expanded
TurnoutExpected moderate turnout in rural areasHigh turnout among Trump base

🧭 Takeaway

Trump's core strength among white working-class voters — especially men, rural voters, and those in the Midwest — remained very strong in 2020, despite polls suggesting weakening support. This group has been a pillar of his electoral coalition, and pollsters have struggled to capture their preferences accurately.

If you're interested, I can also show county-level shifts in working-class regions between 2016 and 2020.

Did more workers vote for Trump in 2016, 2020 and 2024 than for previous Republican presidential candidates?

Yes — more workers voted for Donald Trump in 2016, 2020, and 2024 than for previous Republican presidential candidates, both in absolute numbers and in terms of share of the working-class vote, especially among white, non-college-educated voters.

Here’s a breakdown:


🔹 1. Trump vs. Past GOP Candidates (Absolute Numbers)

Because overall voter turnout has been increasing, it's important to look at both raw vote totals and percentage share of working-class voters.

🗳 Estimated Working-Class Votes for GOP Candidates:

YearGOP CandidateWorking-Class Vote (Est.)Notes
2008John McCain~33–35 millionObama won white working class narrowly in some states
2012Mitt Romney~35–38 millionRomney won 59% of white working class
2016Donald Trump~40–45 millionTrump won 67% of white non-college voters
2020Donald Trump~45–50 millionStrong retention + gains among Latino workers
2024Donald Trump~47–52 million (est.)Continued strength among white workers, growing appeal among others

🔍 These are estimates using:

  • Exit polls

  • Census voter turnout reports

  • Pew and ANES data


🔹 2. Share of the Working-Class Vote

Trump not only got more working-class votes in total due to higher turnout, but also outperformed prior Republicans in percentage terms, especially with:

  • White working-class voters (non-college)

  • Rural and small-town voters

  • Some segments of working-class Latinos and Black men (modest but measurable gains)

📊 White Working-Class Vote (No 4-Year Degree):

YearGOP Share of White Working-Class Vote
2008~58% (McCain)
2012~59% (Romney)
201667% (Trump)
202064% (Trump)
2024~65–70% (Trump) (exit poll est.)

Trump’s 2016 performance was the highest for Republicans in modern history, particularly in the Midwest and Rust Belt.


🔹 3. Why Did More Workers Support Trump Than Past Republicans?

Key Factors:

  • Anti-elite, anti-globalization rhetoric resonated with economically dislocated voters.

  • Trade policy, tariffs, and immigration restrictionism were seen as pro-worker (especially in manufacturing).

  • Cultural messaging on patriotism, religion, guns, and traditional values appealed to white working-class voters.

  • Trump de-emphasized traditional GOP orthodoxy (like entitlement reform) that had alienated many working people under Bush and Romney.

  • He also made gains among non-white working-class voters (Latino men, some Black working-class voters) by 2020 and more so in 2024.


🧾 Summary

Yes, Trump got more working-class votes (especially white non-college voters) than previous GOP candidates — both in numbers and percentage.

📈 Trump expanded the Republican base among workers in a way not seen since Reagan, particularly in the Midwest, Appalachia, and rural areas.

🔄 His support shifted the class balance of the Republican coalition — from a party of business and suburbs to one more dominated by working-class, rural, and non-college voters.


If you're interested, I can also provide comparisons between Trump and Reagan among working-class voters — or explore how union households shifted over the years.

Yes, how did union household shift between pre-Trump Republican candidates and Trump?

Great question. The union household vote has long been a stronghold of the Democratic Party, but Donald Trump made notable inroads, especially compared to earlier Republican candidates. Let’s break down how union household support shifted from pre-Trump Republicans (like McCain and Romney) to Trump in 2016, 2020, and 2024.


🔹 What Is the “Union Household Vote”?

In exit polls, a union household is defined as any household where someone (voter or family member) belongs to a labor union. This includes both public- and private-sector unions.

Historically:

  • Democrats won ~60–70% of union households in most elections since FDR.

  • Republicans consistently struggled with this bloc — until Trump.


📊 Union Household Voting Trends (2000–2024)

YearGOP CandidateGOP % of Union HouseholdsDemocrat %Notes
2000George W. Bush37%62% (Gore)Bush improved slightly from 1996
2004George W. Bush40%59% (Kerry)Higher turnout in Midwest helped Dems
2008John McCain40%59% (Obama)Obama strong in union Midwest
2012Mitt Romney40%58% (Obama)No major GOP gains
2016Donald Trump43–44%~51% (Clinton)Biggest GOP share in decades
2020Donald Trump~45%~52% (Biden)Slight improvement again
2024Donald TrumpEst. 46–48%~50% (Biden)Gains among white and private-sector union members

🔄 What Changed with Trump?

🔺 Republican Gains Under Trump:

  • Trump outperformed McCain and Romney by 3–8 points with union households.

  • He flipped working-class union voters in Rust Belt states (Ohio, Pennsylvania, Michigan).

  • His message resonated with older, white, male union members, especially in construction, manufacturing, and energy.

🔻 Democratic Retention:

  • Democrats still won union households overall, largely due to:

    • Public sector unions (teachers, government employees)

    • Racially diverse unions (e.g., SEIU, hospitality workers)

    • Female union members, who leaned more Democratic


🧠 Key Factors in the Shift

✅ Why Trump Gained:

  • Anti-globalization, anti-NAFTA stance

  • Strong emphasis on manufacturing jobs, coal, and energy independence

  • Messaging on illegal immigration and trade protectionism

  • Appeal to cultural conservatism and “anti-elite” identity

🛑 Why Democrats Still Led:

  • Union leadership and organized campaigns favored Democrats

  • Strong support from public-sector and service unions

  • Continued alignment with labor rights, healthcare, and minimum wage


🧾 Summary

Donald Trump broke decades of Republican underperformance among union households, raising GOP support from ~40% under Bush/Romney to 45–48% by 2024.

His gains were strongest among white, male, private-sector union workers in the Midwest and Rust Belt, and they helped flip key states like Ohio and Pennsylvania.

However, Democrats still held the overall union household majority, especially through support from women, minorities, and public-sector unions.


If you want, I can go deeper into specific unions (e.g., Teamsters, UAW, teachers) or provide county-level examples of heavily unionized areas that swung toward Trump.