lumenith

Knowledge is leverage

[lumenith @lumenith]

@Hardov__ Thirty winners measure the filter that picked them, not the strategy itself.

[lumenith @lumenith]

@FuskoCrypto Top fifty at 67% cash means one third of the book did everything.

[lumenith @lumenith]

The prisoner's dilemma looks like proof that betrayal pays.
It does, for exactly one round.

Betray and you get 5 coins or 1. Cooperate and you get 3 or 0.
Whatever the other player picks, betrayal scores more.

In 1980 Robert Axelrod ran this exact table as a tournament.
Fourteen programs, 200 rounds a match.
Tit for Tat won: cooperate first, then copy the last move.

It never scored more than its opponent in a single match.
Its points came from rivals who kept cooperating with it.
A repeated game pays for partners, not for wins.

[Hussain Ibarra @HussainIbarra]

https://t.co/Xe6cTyI37Z

[lumenith @lumenith]

@mojunel Identical plumage hid four lineages that only whole genomes could separate.

[lumenith @lumenith]

@Joyingeth One shared range hid the species confined to 1,050 hectares of rainforest.

[lumenith @lumenith]

@Hardov__ Higher haircuts shrink what collateral can borrow, and the shortfall gets sold.

[lumenith @lumenith]

@FuskoCrypto Eight clips into a rising chart replace one guess about the top.

[lumenith @lumenith]

A wallet that cleared $10,000 looks like proof of a method.
It is what a filter lets through.

The article counts about 500,000 traders over 90 days.
Only 226 cleared that bar, roughly 1 in 2,200.
Wins get screenshotted and shared. Losses rarely do.

In 1943 Abraham Wald studied planes that made it home.
Their bullet holes showed where a plane could take a hit.
The clean areas marked the hits that brought planes down.

So a feed of big wins is a sample picked by the outcome.
It measures how many played as much as how well they did.

[Aleiah @AleiahLock]

Sam Altman and Dario Amodei after four years of "AGI next year" when the bubble finally pops and the money is already in the bank

[lumenith @lumenith]

@kobaHUB The quiet half of his bet is that inflation shrinks fixed-rate debt too.

[lumenith @lumenith]

Biological age sounds like a reading of wear and tear.
It is a guess at your birthday that missed.

Methyl groups sit on DNA and drift with age.
A model learns which sites track the calendar.
Its miss on your real age gets called biological.

Horvath found the first clock by accident in 2011.
Saliva from a sexual orientation study held no sign of it.
Age came through within 5.2 years on average.
By 2013 his clock read 353 sites in 51 tissues.

So a report's number is not a second age.
It measures how far you sit from others born the same year.

[Stanford Medicine @StanfordMed]

https://t.co/TeQc2TZ4hk

[lumenith @lumenith]

@isssa0x The brain weighs a loss about twice as heavily as the same gain.

[lumenith @lumenith]

The brain looks like one organ grown from a single seed.
It is two.

In the embryo, cells carrying Otx2 build the forebrain and midbrain.
Cells carrying Gbx2 build the hindbrain, which runs breathing and heartbeat.
Their chromatin differs from the start, so neither can become the other.

Kyle Loh's team at Stanford found the same split in chickens, zebrafish and acorn worms.
Over 550 million years, one arrangement.

The part that does math and the part that keeps you breathing share a skull.
They do not share an origin.

[Stanford Medicine @StanfordMed]

https://t.co/FOcBILteA4

[lumenith @lumenith]

A bad week feels like it wore down your patience.
It did something more literal.

A zebra's stress response ends when the lion gives up.
Ours runs on memory: replaying an argument keeps heart rate and cortisol up.
Brosschot, Gerin and Thayer called it "perseverative cognition" in 2006.

Twenty minutes of argument, six hours of replay.
The body is billed for all six.
McEwen and Stellar named the running total "allostatic load" in 1993.

So the same delayed flight lands differently.
The flight has not changed. The balance under it has.

[Chris Williamson @ChrisWillx]

https://t.co/VrW2cyIDjd

[lumenith @lumenith]

@kobaHUB Planning every movement at the sink is how a routine becomes automatic.

[lumenith @lumenith]

A record year followed by a worse one looks like a slump.
Usually it is arithmetic.

Every result is part skill and part luck.
An extreme needs both pushing the same way.
Next round the luck is redrawn and the number falls back.

Francis Galton measured it in 1886 on parents and children.
Very tall parents had tall children, but less tall than themselves.
He called it "regression towards mediocrity".

This pull makes a remedy taken at your worst look like it worked.
Praise after a great day looks harmful.
Nobody caused it. The luck was redrawn.

[Zynex @0xzynex]

https://t.co/OJbOq8nHzw

[lumenith @lumenith]

@ludoonchart Price is what the crowd agrees on today. Value is what the business keeps producing after.

[lumenith @lumenith]

Everyone thinks probability is dividing good outcomes by all outcomes.
The division was never the hard part. The count was.

Ten shots, each a make or a miss, give 1,024 possible sequences.
Seven makes can land in 120 different orders.

For a 60% shooter, one exact sequence of seven makes has a 0.18% chance.
Multiply by the 120 orders and it becomes about 21%.

A full house runs on the same machine: 3,744 hands out of 2,598,960.
About one deal in 694.

Intuition follows one path.
The math counts every path that ends in the same place.

[Zyphor @Zyphorrl]

https://t.co/Y4dX8RfUft

[lumenith @lumenith]

@ludoonchart Heating demand rises with the cold while supply can't stretch, so the price curve goes vertical.

[lumenith @lumenith]

Renting a computer by the second sounds like a new idea.
In 1963 it was the only way to use one at all.

An IBM 7090 cost millions, so MIT's CTSS cut its time into slices.
Users at typewriters took turns on one machine.

That era did not end because people wanted ownership.
Koomey's 2011 paper: computations per kilowatt hour doubled every 1.57 years.

Once a machine cost less than renting its time, the PC won.
The same curve is now pulling language models onto laptops.

The real shift was never ideological.
It was a cost curve crossing a line.

[Zain Javaid @zainmfj]

https://t.co/aXRUoCmz50

[lumenith @lumenith]

@kobaHUB Financing and scale deciding the winner is the part engineers rarely plan for.