Equations and E. Coli data are never going to be as convincing as simulating the whole genomes with all the chromosomes and non-coding regions of a population of thousands of people over thousands of generations, with overlapping generations, population splits, migration, mutation hotspot map, sexual selection, and all the other wrinkles that give deniers refuges to deny less fully-realistic models. The software and the setup to realistically sumulate human evolution are now available [link below] open-source and tested, decades of work not required, just hours to days of cloud compute. There are settings and initial conditions to the existing sims that you could make more realistic, which may bring the quanitative results closer to the equations, but the genetic sims already demonstrate your point qualitatively: the conventional evolutionary theory requires far too many generations to work and the few fixed mutations are hardly ever advantageous.
From a note:
The software for realistic whole-genome simulations of the whole human population over thousands of generations only became available within the last two years. [more like 18 mo.] Early results show that tens of thousands of “burn-in” generations have to be added to simulations before the migration out of Africa to get the gene fixation rates expected by the conventional simplified evolutionary model.
“SimHumanity: Using SLiM 5.0 to run whole-genome simulations of human evolution” Haller et al. Human Population Genetics and Genomics 2025;5(4):0006
[quote]
[model 3, a model running about 2k generations, including African and Asian population splits, with genomic tree records added to model 2, positive and negative fitness mutations explicitly simulated and neutral mutations from the first~77k generations of model 2 added back after the fact] took 16472 seconds (4.58 hours) and used 78.0 GB of memory, running on the same test machine. At the end of the run there were 620,596 segregating mutations (99.99% of which were deleterious, 0.0135% beneficial) and 85 fixations (29.41% of which were deleterious, 70.59% beneficial). By comparison, model 2 above had 625,065 segregating mutations (99.98% deleterious, 0.0152% beneficial) and 6028 fixations (94.77% deleterious, 5.22% beneficial). From this comparison it can be seen that model 3, due to its curtailed burn-in period, unsurprisingly had far fewer fixations overall
Uh oh - I think you ruined your own argument by mistake. How else could a sloth be made to play bass like that? I wish my hands could do that AND be shaped perfectly for it.. Had to be evolution - no other way.
The above approach to working with knowledge is probably going to be the best thing to ever come out of AI for humanity. MITTENS is the obvious proof of concept of just how effective and revolutionary using AI in this way can be. The best study on how to use AI this way is the bestseller Hardcoded: AI and the End of the Scientific Consensus, although the main purpose of the book (which it expertly fulfills) is to be "the definitive account of how AI systems trained on the corrupted corpus of modern science have inherited every pathology of the institutions that produced them". It's the third book in the trilogy that is currently turning the life sciences on their head.
And by the way, Vox, you make some really awesome AI-augmented grooves for being one of the world's top truckers!
The scientific evidence is in, the numbers are solid, and the proof is ironclad and mathematically irrefutable - Vox has severely overdosed on too much Prince, to the point that no recovery is possible.
Although it's debatable that one would WANT to recover...
The monkey playing that particular instrument is a subtle but brilliant metaphor for the fact that the Keytar was a rare beneficial mutation in the evolution of music.
Equations and E. Coli data are never going to be as convincing as simulating the whole genomes with all the chromosomes and non-coding regions of a population of thousands of people over thousands of generations, with overlapping generations, population splits, migration, mutation hotspot map, sexual selection, and all the other wrinkles that give deniers refuges to deny less fully-realistic models. The software and the setup to realistically sumulate human evolution are now available [link below] open-source and tested, decades of work not required, just hours to days of cloud compute. There are settings and initial conditions to the existing sims that you could make more realistic, which may bring the quanitative results closer to the equations, but the genetic sims already demonstrate your point qualitatively: the conventional evolutionary theory requires far too many generations to work and the few fixed mutations are hardly ever advantageous.
From a note:
The software for realistic whole-genome simulations of the whole human population over thousands of generations only became available within the last two years. [more like 18 mo.] Early results show that tens of thousands of “burn-in” generations have to be added to simulations before the migration out of Africa to get the gene fixation rates expected by the conventional simplified evolutionary model.
“SimHumanity: Using SLiM 5.0 to run whole-genome simulations of human evolution” Haller et al. Human Population Genetics and Genomics 2025;5(4):0006
[quote]
[model 3, a model running about 2k generations, including African and Asian population splits, with genomic tree records added to model 2, positive and negative fitness mutations explicitly simulated and neutral mutations from the first~77k generations of model 2 added back after the fact] took 16472 seconds (4.58 hours) and used 78.0 GB of memory, running on the same test machine. At the end of the run there were 620,596 segregating mutations (99.99% of which were deleterious, 0.0135% beneficial) and 85 fixations (29.41% of which were deleterious, 70.59% beneficial). By comparison, model 2 above had 625,065 segregating mutations (99.98% deleterious, 0.0152% beneficial) and 6028 fixations (94.77% deleterious, 5.22% beneficial). From this comparison it can be seen that model 3, due to its curtailed burn-in period, unsurprisingly had far fewer fixations overall
[end quote]
https://www.pivotscipub.com/hpgg/5/4/0006/html
Feelin ready to educate others with your newly found Ph.D? Want to let the High-C ring with the Fox?
With the lyrics, you can!
--- Darwin is Dead, but the Funk's Begun ---
Got my mind on those genes baby! Fixation's slow.
Hundreds of generations, don't you know?
Natural selection's got its groove by its crawl.
Evolution's math some mighty tall wall.
Mutations happen, they take their sweet time.
Fixation in the genes, it's a mountain to climb.
Survival of the fittest, you heard it before.
But it's the math that's tricky.
There's so much more!
Get your groove on,
cause Darwin's dead yo!
Math don't lie and that's a fact!
Generations passing, one by one.
Fixation's slow, but the funk's begun!
DNA is twisted, a double helix spin-(spiral!).
The changes in the code, man they take years to win.
Thousands of years just to make a modest shift.
Mathematical reality disproves genetic drift.
Chance has been disproven, the numbers don't compute.
Mutation nor fixation means the question is moot.
Populations change, but it's super slow.
It's a long-winding journey through genetic-flow!
So we DNA dance, shake it to the groove.
Gotta feel that rhythm, there is nothing to prove.
Natural selection's not a night at a club.
But we are caught in this math,
there is the rub!
Get your groove on,
cause Darwin's dead yo!
Math don't lie and that's a fact!
Generations passing, one by one.
Fixation's slow, but the funk's begun!
Get your groove on,
cause Darwin's dead yo!
Math don't lie and that's a fact!
Generations passing, one by one.
Fixation's slow, but the funk's begun!
(GUITAR SOLO)
Get your groove on,
cause Darwin's dead yo!
Math don't lie and that's a fact!
Generations passing, you comprehend.
Fixation's slow, but the funk don't end!
The comment section does not allow adding paragraph breaks, you may read lyrics with proper paragraph breaks at https://darklordfan963679.substack.com/p/lyrics-to-the-intelligent-groove
"Darwin, we have a problem." - Vox Day
The evsnescent lemur tail, appearing snd vanishing from guitarists, shows why fixation is ficjle.
But the Funk prevails!
Uh oh - I think you ruined your own argument by mistake. How else could a sloth be made to play bass like that? I wish my hands could do that AND be shaped perfectly for it.. Had to be evolution - no other way.
Don’t even get me started on the foxy FOX 🦊.
Checkmate VD.
With apologies to Rudyard Kipling…
Dodging the Question
Vox Day lays his case on the table,
And Myers examines his hat:
“His schooling is hardly sufficient;
We needn’t look further than that.”
And that is called dodging the question,
And the question survives all the same;
For a flaw in the man’s education
Is no proof of a flaw in his claim.
“He’s used a machine in his writing!
His prose has the taint of AI!”
Then show where the premises falter,
And show where the numbers go awry.
For that is still dodging the question,
Though the audience cheers at the name;
You may curse every tool in the workshop,
But the sum will come out just the same.
You may laugh at Probability Zero,
You may sneer at the fellow who wrote;
But a sneer is no counterexample,
And a proof is no matter of vote.
And that is called dodging the question,
With a jeer where an answer should stand;
You have battered the name on the cover
While the argument waits in your hand.
If the odds have been wrongly computed,
If the model has smuggled a choice,
Then expose it in plain mathematics;
There is no need to raise up your voice.
For the cure for a faulty deduction
Is to show where deduction went wrong;
And the louder you shout at the author,
The longer the question lives on.
We ask for no faith in the author,
No bow to his banner or creed;
We ask you to point to the error
In the argument offered to read.
For once you make dodging the question
The test by which arguments fall,
You may win every cheer in the lecture
And have answered no question at all.
Sorry, Mr. Vox Day. The genomic explanation is the fun part.
Disco. FUNK. I ask you-!
I always wanted a PhD in Evolutionary Biology, thanks!
...And just think of the money I saved!
i find IFLScience's lack of guitar solo disturbing.
The above approach to working with knowledge is probably going to be the best thing to ever come out of AI for humanity. MITTENS is the obvious proof of concept of just how effective and revolutionary using AI in this way can be. The best study on how to use AI this way is the bestseller Hardcoded: AI and the End of the Scientific Consensus, although the main purpose of the book (which it expertly fulfills) is to be "the definitive account of how AI systems trained on the corrupted corpus of modern science have inherited every pathology of the institutions that produced them". It's the third book in the trilogy that is currently turning the life sciences on their head.
And by the way, Vox, you make some really awesome AI-augmented grooves for being one of the world's top truckers!
The scientific evidence is in, the numbers are solid, and the proof is ironclad and mathematically irrefutable - Vox has severely overdosed on too much Prince, to the point that no recovery is possible.
Although it's debatable that one would WANT to recover...
Bass is in the blood, baby!
I grew up in the Minneapolis Sound.
The monkey playing that particular instrument is a subtle but brilliant metaphor for the fact that the Keytar was a rare beneficial mutation in the evolution of music.
https://www.facebook.com/NotAnExactScienceShow/photos/keytar-platypus-venn-diagram-by-tenso-graphicswwwthreadlesscomtensovennsday/10152588503450390/
i will one day get a keytar for my live performances if it kills all of us. (might wanna write that down).
This is not negiotiable. Do it.
This is art!
Holy crap, he's LITERALLY dancing on evolution's grave.
That guitar solo reminded me to pour another one out for John. Solid!
Baby? If you ain't dancing through this song & working your groove, you ain't got no chili.
"Toast!"
hahahahahaha