Minute of Silence
On Feb. 11, Ukraine's parliament passed Bill No. 14144, formally establishing a daily nationwide moment of silence at 9 a.m. Kyiv time.
To show respect towards this decision and to keep alive memory of all those murdered, killed, tortured, injured &c. for unjust reasons during all senseless wars and atrocities of the past, present, and future, the counter for minute of silence is hereby activated in this room.
Thanks in advance for not disturbing this contemplative process.
Klasse Leben
WiSeAE56
Short History of the Learning Machine
Between Archetypes & Myths: Machine as a Projection of the Human Self
Between Archetypes & Myths
Machine as a projection of the human self
Faciamus machinam ad imaginem et similitudinem nostram.
Let us make the machine in our image and likeness.
Before the Machine
Long before humans could build autonomous machines, they could imagine them.
Artificial humans. Servants. Lovers. Guardians. Weapons. Oracles. Monsters.
The technological history of machines is preceded by a much older imaginary history of artificial agency.
An Archetype Is Not a Blueprint
Myths do not predict robots, computers or artificial intelligence.
But they repeatedly stage similar relationships between humans and artificial beings.
Who creates whom?
Who commands whom?
What should an artificial being know?
What happens when it acts for itself?
Machine as Projection
We rarely imagine machines neutrally.
We project into them our bodies.
Our desires.
Our labor.
Our violence.
Our knowledge.
Our fear of losing control.
What Is a Machine?
Before machines became motors, engines and computers, the word already meant something broader:
an ingenious means by which something becomes possible.
μηχανή → machina → machine
Greek μηχανή — mēkhanē: instrument, device, contrivance, means.
Latin machina: machine, engine, apparatus, contrivance.
Middle French machine → English machine.
The semantic field is older and broader than today's image of gears, motors and electronics.
Deus ex Machina
In Greek theatre, a mēkhanē could be the stage apparatus that lifted a god into view.
Hence: deus ex machina — the god from the machine.
human construction → superhuman appearance
The machine makes visible powers that exceed the ordinary human body.
Technē
The machine begins not only with matter, but with know-how.
Greek technē means skilled, rule-governed making.
thought → technē → artifact
Before artificial agency comes the intelligence of the maker.
The Divine Workshop
Greek mythology gives us an extraordinary laboratory of artificial beings.
Athena represents intelligent making.
Hephaestus represents the workshop in which artifacts begin to act.
Athena: Intelligence of Making
Athena is associated with craft, weaving, strategy and intelligent production.
She is not primarily the goddess of machines.
She represents something more fundamental:
the intelligence embodied in making.
Hephaestus: When Artifacts Act
With Hephaestus, myth moves from intelligent making toward artificial agency.
His workshop contains self-moving devices and artificial attendants.
The question changes:
Can we make an object?
becomes
Can we make an object that acts?
Machine as Human
In the Iliad, Hephaestus is assisted by artificial golden maidens endowed with capacities associated with living persons.
The ancient fantasy is already recognizably humanoid:
Can the made thing become sufficiently like its maker?
Machine as Servant
One of the most persistent fantasies is delegation.
The artificial being carries, assists, manufactures, obeys.
labor → delegated labor → autonomous labor
The machine performs what we cannot, will not, or no longer need to do ourselves.
Machine as Prosthesis
A machine need not imitate the whole human.
It may instead extend a capacity.
lever → force
vehicle → movement
writing → memory
computer → calculation
artificial agent → cognition?
Yan Shi: The Artificial Human
The Chinese Liezi tells of the artificer Yan Shi presenting King Mu with a human-like artificial performer.
It walks. It sings. It performs.
When its behavior becomes disturbingly lifelike, the artificer takes it apart.
Open the Black Box
apparently human → disassembly → mechanism
The story stages a remarkably durable question:
What remains of the human when we open the box?
Machines in Our Image
Artificial beings are not merely tools.
They become surfaces onto which humans project desire, identity and self-understanding.
Machine as Lover
Pygmalion creates an ideal human form and falls in love with his creation.
Venus / Aphrodite grants the statue life.
The question is not merely:
Can we reproduce a human?
but:
Can we manufacture the human we desire?
Machine as Mirror
Artificial beings force us to ask what supposedly distinguishes us.
If it moves like us?
If it speaks?
If it remembers?
If it desires?
If it creates?
If it learns?
By constructing the non-human, we investigate the human.
Machine as Double
The artificial human is also a duplicate.
A second body.
A copy that may become difficult to distinguish from an original.
The double turns technology into an identity problem:
How do I know that you are one of us?
Machines at the Threshold
Some artificial beings do not imitate us primarily as workers or companions.
They stand between inside and outside.
They decide who may approach, enter or survive.
Machine as Protector
Talos is the bronze guardian of Crete.
He patrols the island and prevents outsiders from landing.
human guard → artificial guard → tireless guard
Protection has been delegated to an artificial body.
Machine as Weapon
Talos also exposes the ambiguity of protection.
The same autonomous force that protects one community threatens whoever approaches it.
protector for us = weapon against them
Who gives the command?
Who defines the enemy?
Guardian of the Threshold
The guardian archetype is broader than the machine.
Cherubim guard the Tree of Life.
The Sphinx guards passage through a riddle.
Talos guards an island.
Artificial watchdogs guard a palace.
boundary + test + permission
The Sphinx and the CAPTCHA
The Sphinx:
Prove that you know before I let you pass.
The CAPTCHA:
Prove that you are human before I let you pass.
When machines learn to pass the test, the ancient threshold problem returns in inverted form.
Machine as Oracle
A machine can do more than move.
It can contain knowledge unavailable to its operators.
The mythical Argo gives us an extraordinary example.
Athena + Argus + Dodona → Argo
Argus builds the ship.
Athena contributes divine technē.
She installs a special beam from the sacred oak of Dodona, a site of divination.
The resulting architecture can be described, anachronistically but suggestively, as:
vehicle + embedded oracle
When the Argo Speaks
In the Argonautica, the Argo itself cries out and urges the heroes to depart.
Elsewhere the sacred beam communicates information and guidance.
The artifact possesses knowledge that its human operators do not.
world → hidden information → artifact → human action
An Intelligent Artifact?
The Argo is more than a passive object.
It speaks.
It advises.
It carries privileged information.
It is therefore a remarkably early imagination of intelligence embedded in an artifact.
But the Argo Does Not Learn
The distinction is crucial.
The Argo possesses knowledge.
But its defining capacities are installed, not acquired through experience.
A machine containing knowledge is not necessarily a learning machine.
The Word Becomes Agent
The Golem introduces another extraordinary structure:
matter + symbolic information → agent
Machine as Programmed Being
In Jewish traditions, an artificial human may be formed from inorganic matter through esoteric knowledge.
Later popular versions emphasize sacred names, letters or words associated with animation and control.
The physical body alone is insufficient.
Information activates matter.
A Historical Warning: Prague
Myths themselves have histories.
The famous association between Rabbi Loew, Prague and a protective Golem is much later than Rabbi Loew himself.
The familiar Prague narrative develops gradually and becomes especially prominent in modern literary culture.
We must ask not only what a myth says, but when people began telling it that way.
Machine as Uncontrolled Creation
The servant is useful only while its behavior remains inside the creator's intentions.
creation → autonomy → unpredictability → fear
The question becomes:
Can the maker remain master of the made?
Machine as Monster
The monster violates categories:
object / subject
dead / alive
natural / artificial
tool / agent
controlled / autonomous
The frightening machine makes the boundary itself unstable.
A Mythological Bestiary of Machines
Human — Can we manufacture ourselves?
Servant — Can we delegate labor?
Prosthesis — Can we extend ourselves?
Lover — Can we manufacture intimacy?
Protector — Can we delegate protection?
Weapon — Can artificial agency project violence?
Guardian — Can it decide who may pass?
Oracle — Can it know what we do not?
Mirror — What does it reveal about us?
Monster — What if it escapes our intentions?
One Archetype Is Still Missing
The golden maidens can act.
Talos can guard.
Pygmalion's statue can live.
The Golem can obey.
The Argo can advise.
But their defining abilities are generally given to them.
What happens when the artificial being can change itself through experience?
Machine as Learner.
What Is a Learning Machine?
A traditional machine is built in a particular state and executes a structure provided by its maker.
A learning system differs in one decisive respect:
its encounter with the world changes what it subsequently does.
experience → internal change → changed future behavior
Two Histories
LEARNING
myth
initiation
memory
imitation
association
conditioning
adaptation
MACHINE
myth
technē
automata
mechanism
calculation
programmability
computation
LEARNING + MACHINE
The Persistent Question
The history of machines is also the history of what humans have wished to externalize.
strength
movement
memory
judgment
companionship
agency
learning
In Our Image
We first imagined machines as versions of ourselves.
Then we built machines that extended parts of ourselves.
Eventually we built machines capable of modifying themselves through experience.
The artificial being is ancient.
The learning machine is the historical novelty.
Discussion
Which archetype best describes your smartphone?
Which describes a large language model?
Which describes an autonomous vehicle?
Which describes social media?
Which describes the machines you imagine for the future?
And have genuinely new archetypes appeared?
Textual Anchors
Primary traditions worth returning to: Homer, Iliad XVIII; Apollonius Rhodius, Argonautica; Ovid, Metamorphoses X; the received Liezi; Jewish Golem traditions.
Golem
Argo: speaking Dodonian timber embedded in Athena's ship.
Talos: bronze guardian of Crete.
Pygmalion: desired artificial human brought to life.
Golem: artificial human animated through knowledge, letters or sacred language in evolving Jewish traditions.
Alles ist Zahl!
Pythagoras's Claim
“All is number.” (πάντα αριθμός) — attributed to Pythagoras, 6th century BCE.
For Pythagoras and his followers, numbers were not just a tool for counting. They were the hidden structure of reality itself — the ratios that govern music, geometry, even the harmony of the cosmos.
Today we trace that claim all the way to the silicon running inside your phone.
Why It Mattered
Here is the thread we will follow: a number can be written down in many different ways — and the way you write it changes what you can do with it.
That was true for Roman numerals. It is just as true for the floating-point formats inside a modern AI chip.
Counting Before Zero
Rome counted like this: I, V, X, L, C, D, M.
No zero. No place value — the position of a symbol barely matters, only which symbols appear and in what combination.
Try multiplying MDCLXVI by XLIV in your head. Now try 1666 × 44. The second one is not just easier — it is a genuinely different kind of thinking.
An Older Theorem
Centuries before Pythagoras, the Baudhayana Sulba Sutra (India, c. 800 BCE) already states the same relationship between the sides of a right triangle.
The mathematics was not a single European discovery — it was discovered, independently and repeatedly, wherever people needed to measure, build, and trade.
The Missing Symbol
Every numeral system so far shares the same gap: no symbol for nothing.
Without a zero, there is no clean way to write a number as a position-dependent code — and without position, arithmetic stays a craft for specialists, not a procedure anyone can follow.
The Invention of Zero
Zero emerges independently more than once: among the Maya of Mesoamerica, and in India, where the mathematician Brahmagupta (7th century CE) is the first to give it formal rules of arithmetic.
It reaches the Islamic world through scholars like Muhammad ibn Musa al-Khwarizmi (9th century), whose own name — Latinized as algorismus — gives us the word algorithm.
A number system and a method for calculating with it arrive together. That is not a coincidence.
Zero Reaches Europe
In 1202, Leonardo of Pisa — better known as Fibonacci — publishes Liber Abaci, introducing Hindu-Arabic numerals (zero included) to European merchants.
The pitch was practical: this notation makes bookkeeping and trade dramatically easier than the Roman system merchants were already using.
Positional Notation
Here is the real leap: the same ten symbols, re-used, mean something different depending on where they sit.
106 and 160 use identical digits. Position alone carries the meaning.
This single idea — a small, finite alphabet whose meaning depends on position — is the one that modern computing was built on top of.
Leibniz's Dream
Gottfried Wilhelm Leibniz takes positional notation somewhere further: in 1703, he formally describes arithmetic in base 2 — binary.
He was reportedly delighted to find the same two-symbol logic echoed in the hexagrams of the I Ching, and saw in binary something close to a theological image: everything created from nothing (0) and unity (1).
Three centuries later, his once-mystical number system runs every computer on Earth.
Why Base Two?
Binary is just another positional system — base 2 instead of base 10. Only two symbols: 0 and 1.
That turns out to be the whole point. Building a reliable physical device with ten stable, distinguishable states is hard. Building one with two — voltage on, voltage off — is comparatively easy, cheap, and extremely reliable at enormous scale.
Every transistor in every chip is, at heart, a two-state switch pretending to be one of Leibniz's digits.
From Numbers to Logic
George Boole (1854) notices that 0 and 1 can represent more than quantity — they can represent truth: false and true.
Boolean algebra — AND, OR, NOT — becomes the logical skeleton underneath every circuit in every digital machine.
The same two symbols now carry both arithmetic and logic. That overlap is why a computer can calculate and also decide.
Bits, Bytes, and Beyond
One binary digit is a bit. Group eight of them and you get a byte — 256 possible values.
Hexadecimal (base 16) exists purely as a convenience: one hex digit covers exactly 4 bits, so two hex digits describe a whole byte cleanly — far easier for a human to read than a string of eight zeros and ones.
You already used one today, probably without noticing: a colour code, a MAC address, a cryptographic key — all hexadecimal shorthand for binary underneath.
A New Question
So far: how a culture chooses to write a number. Now: how a machine chooses to store one — and why, in a modern AI system, that choice is no longer a historical curiosity.
It is an engineering decision with real, measurable consequences.
Integers vs Floats
Integer formats store whole numbers exactly. int8: one byte, 256 distinct values — typically -128 to 127.
Floating-point formats trade some of that exactness for enormous range: a sign bit, an exponent, and a mantissa (the significant digits) — the same idea as scientific notation, just in binary.
More bits generally buys you more precision, more range, or both. Fewer bits buys you speed and memory — at a cost.
f16 vs bf16
Both fp16 and bf16 use 16 bits total — half the size of ordinary fp32 — but split those bits differently.
fp16 (IEEE half precision): 1 sign + 5 exponent + 10 mantissa bits. More precision, a narrower range — values that are too large or too small simply overflow to infinity or zero.
bf16 ("brain float", from Google): 1 sign + 8 exponent + 7 mantissa bits. The same exponent range as full 32-bit floats, less precision — far more forgiving during the unstable arithmetic of training a neural network.
Same bit budget. Different trade-off between "how exact" and "how big."
Why int8 for Inference
Models are usually trained in fp16, bf16, or fp32 — the precision that stability during training needs.
Once trained, they are often quantized down to int8 for inference (actually running the model): roughly a quarter of the memory of fp32, and hardware built for int8 can push through far more operations per second than the same silicon running floating point.
The accuracy cost is usually small. The speed and energy savings, at the scale a real deployment runs at, are not.
Case Study: NVIDIA Jetson Orin
Take a real, current piece of edge-AI hardware: NVIDIA's Jetson AGX Orin (64GB module).
NVIDIA's own published figure: up to 275 TOPS (trillion operations per second) running INT8, with structured sparsity.
Switch the same chip to FP16 / BF16 and achievable throughput drops — roughly by half, since each floating-point operation occupies more of the tensor core's datapath than an 8-bit integer one.
Same transistors. Same chip. The number format you choose to compute in changes your real-world throughput by a factor of two or more.
Closing: Same Question, New Stakes
Pythagoras claimed number was the hidden structure of reality.
Three thousand years later, the question "what exactly is this number, and how many bits does it cost to represent it" is no longer philosophy — it decides how fast your model runs, how much it costs to train, and how much energy it burns doing so.
The history of numerals is, in a very literal sense, the history of computation.
Unplugged #0
Cogs, Steam, Algorithm
Alan and his Enigma
Claude, Marvin, John &c.
Machine Learning #0
The Latent Space
Unplugged #1
The Omega Point
Language models
Artificial Agents
Central Problem of AI ethics
Human-Machine Peer Learning
Machines and humans can learn with each other
Memes, Mindworms, Memplexes &c.
Memes in Alltag
Session 0 · Memetics begins at home.
Before Internet memes, before ideology, before theory: gestures, tunes, habits, rituals, phrases, drawings, interfaces and everyday acts of copying.
Meme
“A meme is whatever is copied from person to person.”
— Susan Blackmore, working definition
For today, let us start with this deliberately simple definition.
Copied?
Copying does not necessarily mean making an exact duplicate.
You hear a joke.
You remember the structure, forget several words, change a character...
...and tell it to somebody else.
Something survived the transformation.
Two Worlds
external expression → perception → internal reconstruction → expression
A meme can appear outside us: as sound, gesture, image, text, object, ritual.
It can be reconstructed inside us: as memory, rule, association, expectation, intention.
Then it can become external again.
What exactly is being copied?
A Word of Warning
Memetics is a theory among other theories.
Human culture can also be studied through psychology, anthropology, sociology, cultural evolution, semiotics, media theory, neuroscience and many other perspectives.
Memetics highlights replication, variation and selection.
It does not explain the whole human being.
And although it may begin to look as if everything is a meme...
...it is not.
Not Everything That Spreads Is a Meme
A virus spreads. That does not make it a cultural meme.
Genes are copied. That does not make them memes.
You discover by yourself that fire is hot. No person-to-person copying was necessary.
A yawn may trigger another yawn through an evolved response rather than culturally learned imitation.
Spread alone is not enough.
Where Is the Meme?
Is the meme the sentence?
The idea expressed by the sentence?
The memory in somebody's brain?
The rule that allows the sentence to be reconstructed?
The pattern distributed across people, media and artifacts?
Different memetic theories answer differently.
Memethical Engineering
Memethical engineering is the deliberate design, modification, and circulation of memes and memetic environments in accordance with explicit ethical goals.
For example: counteracting propaganda and manipulation.
Increasing epistemic and cultural diversity.
Supporting critical reflection.
Increasing resilience against harmful or monopolizing meme complexes.
Engineering the Environment
Memetic engineering asks: How can this spread?
Memethical engineering must additionally ask: Should it spread?
Who benefits?
What disappears when it succeeds?
Can people recognize the mechanism?
Can alternatives survive?
Success ≠ virality.
Memes in Alltag
Let us now temporarily leave theory behind.
Look at ordinary life as a landscape of things passing from person to person.
We will move from the smallest and fastest patterns...
...toward increasingly complex cultural structures.
1 · Contagion
😮 yawning
😂 laughter (e.g. Tanganyika laughter epidemic of 1962)
👏 applause / standing ovation
👀 looking where everybody else is looking
🌊 the stadium wave
These are useful starting examples because copying is immediately visible.
But some may be biological contagion rather than memes in Blackmore's strict sense.
Experiment
One person suddenly looks intensely at the ceiling.
How many people look too?
Did information pass?
Was anything learned?
Would you call it a meme?
2 · Gestures
🤝 handshake
✋ high five
👊 fist bump
✌ peace sign
🫶 hand-heart
air quotes
😘 kiss (hand ? cheek ? lips ?)
nodding (vertical ? horizontal ?)
A tiny movement can carry socially learned meaning.
Gesture = Movement + Convention
The physical movement alone is not enough.
A raised thumb means something because other people already know how to interpret it.
The same movement can mean different things in different cultures.
A meme therefore often contains both performance and interpretation.
3 · Rituals
🎂 birthday cake + candles + song + wish
🥂 touching glasses before drinking
🎆 counting down at New Year
🪵 knocking on wood
🤞 crossing fingers
💰 throwing a coin into a fountain
Memes can become coordinated social scripts.
The Queue
At the highest floor of the Eiffel Tower, there are two elevators facing each other.
If human beings were fully rational beings, two equally long queues should have formed, each unfolding in its own direction.
But again and again, only one queue appears.
A long line of visitors patiently waits for one elevator...
...while the other elevator stays empty and available.
People tend to infer their own action not from the total situation, but from the behavior of the people already in front of them.
The Elevator
Enter.
Turn around.
Face the door.
Avoid prolonged eye contact.
Become strangely silent.
Who taught us this?
Many everyday conventions are acquired without an explicit teacher.
4 · Words
“Actually...”
“However...”
“You know?”
“At the end of the day...”
euh… [øː] , bah oui
блядь, f*ck, vole ...
Language is full of units which seem to "live their own life".
Catchphrases
A phrase can detach itself from its original situation.
It becomes reusable.
People recognize it.
They repeat it because others recognize it.
Recognition itself can increase its probability of further replication.
Familiarity can become its own selection advantage.
5 · Earworms
A melody enters through the ears.
Hours later it is still playing internally.
You hum it.
Somebody else hears it.
Now it may inhabit another mind.
An earworm is almost a cartoonishly clear memetic lifecycle.
Sound Memes
nursery rhymes
football chants
advertising jingles
melodies with changing lyrics
TikTok audio fragments detached from their original context
The carrier can change while the recognizable pattern survives.
6 · Drawings
♥ a heart that looks nothing like an anatomical heart
★ a five-pointed star
☀ a circle with rays
🏠 a square house with a triangular roof
💡 a light bulb meaning “idea”
We learn visual conventions by repeatedly encountering them.
The Cool S
Who knows how to draw the “Cool S”?
Who taught you?
Where did you first see it?
Do you know who invented it?
Why did generations of schoolchildren reproduce it?
A near-perfect classroom specimen of anonymous cultural transmission.
7 · Interface Memes
💾 floppy disk = save
🔍 magnifying glass = search
⚙ gear = settings
🗑 bin = delete
♥ heart = like
☰ three lines = menu
The original object can disappear while its symbolic descendant survives.
Memes Can Outlive Their World
Many people click a floppy disk icon without ever having saved a file onto a floppy disk.
The symbol has become detached from the technology that produced it.
Its survival depends on continued copying inside interface culture.
Cultural ancestry can remain visible after functional ancestry has vanished.
8 · Fashion
ripped jeans
mullets
moustaches
Crocs
rolling trouser legs
particular sneaker styles
black clothing in particular creative milieus
Bodies become media.
Fashion Evolves
Someone adopts a style.
Others notice it.
Some imitate it.
Some exaggerate it.
Some reject it.
Some revive an older version thirty years later.
Replication produces variation; variation produces selection.
9 · Food & Household Knowledge
family recipes
sourdough starters
cutting sandwiches diagonally
particular ways of folding pizza
“five-second rule”
putting a wet phone into rice
Useful, useless and questionable practices can all be culturally transmitted.
Transmission ≠ Truth
A meme does not have to be true to spread.
It does not have to be useful.
It does not have to benefit its host.
It merely needs properties that allow further transmission.
This is one of the most important—and most uncomfortable—ideas in memetics.
10 · Social Formats
the elevator pitch
brainstorming with Post-its
the agenda slide
the TED-style talk
“three things I learned...”
the Venn diagram
the iceberg diagram
Sometimes the meme is not the content but the format.
11 · Internet Memes
image macros
reaction GIFs
“POV:”
“starter pack”
“nobody: / absolutely nobody:”
“tell me X without telling me X”
“how it started / how it's going”
The Internet did not invent memes. It created an extraordinarily fast memetic environment.
Template + Variation
Internet memes make one property especially visible:
a stable template + many mutations.
Distracted Boyfriend.
Drake Hotline Bling.
Woman Yelling at a Cat.
The image remains recognizable while captions and meanings mutate.
Variation is not failure of copying. It is part of the evolutionary process.
12 · Stories & Rumours
urban legends
schoolyard rumours
ghost stories
“a friend of a friend once...”
life hacks
folk explanations
Stories are compressed, reconstructed and altered every time they move.
13 · Beliefs
superstitions
stereotypes
ideas about success
ideas about beauty
ideas about what a “proper career” looks like
ideas about how families, schools or workplaces should function
At this scale, memes begin to shape perception itself.
From Meme to Memeplex
Memes rarely live alone.
They support, trigger and protect one another.
A fashion includes clothes, vocabulary, gestures, music, places and identities.
A profession includes practices, jargon, values, rituals and status symbols.
Religions, ideologies, artistic movements and institutions can contain enormous interconnected systems of culturally transmitted elements.
Such complexes are often called memeplexes.
Replication
Something is encountered.
Something is reconstructed.
Something is expressed again.
Another person encounters it.
Replication does not require perfect duplication.
Variation
Misremember it.
Translate it.
Improve it.
Simplify it.
Exaggerate it.
Parody it.
Combine it with another meme.
Every transmission is an opportunity for mutation.
Selection
Some variants disappear.
Some are remembered.
Some are repeated.
Some fit a particular medium particularly well.
Some become part of institutions and persist for centuries.
Cultural environments select.
What Makes a Meme Successful?
easy to remember?
easy to reproduce?
emotionally intense?
funny?
useful?
beautiful?
shocking?
socially rewarded?
already familiar?
And are these the same properties that make a meme good?
The Memethical Question
A communication system optimized only for replication may reward whatever replicates best.
But the fastest-spreading cultural environment is not necessarily the most desirable one.
Memethical engineering asks how environments might instead preserve:
diversity · autonomy · transparency · reflection · plurality · resilience
Exercise · Find One
Think of one meme you encountered today.
Not necessarily an Internet meme.
Where did you encounter it?
What was its external form?
What did you reconstruct internally?
Did you reproduce it?
Did you modify it?
Why might this particular pattern survive?
Exercise · Classify It
gesture?
ritual?
phrase?
melody?
visual convention?
interface convention?
fashion?
social script?
story or belief?
memeplex?
Or perhaps: not a meme at all?
Three Questions to Keep
1. What is being copied?
2. Why does this variant survive?
3. What kind of cultural environment does its survival create?
Memes in Alltag
A handshake.
A birthday song.
A stupid little drawing in a school notebook.
A phrase you suddenly notice yourself saying.
A floppy disk that survived its floppy disk.
A belief nobody remembers learning.
Culture replicates through us—but never without being reconstructed by us.
The Second Replicator
Before Darwin
For most of human history, the diversity of life was explained by one simple idea: species are fixed, created once, unchanging.
A rabbit begets a rabbit. An oak begets an oak. World without end.
Even naturalists who classified thousands of species rarely asked: where did all this variety come from?
Darwin's Insight
In 1859, Charles Darwin proposed something radical: species are not fixed.
They change, slowly, over generations — new species arise from old ones through an entirely natural, unplanned process.
He called it natural selection.
No designer. No goal. Just a mechanism.
The Engine of Evolution
Replication + Variation + Selection = Evolution
Darwin never used these three words this way — but this is the abstract engine hiding inside his argument.
Three ingredients. Run them together, over and over, and something astonishing happens: design without a designer.
Replication
Something must be copied from one generation to the next.
Offspring resemble parents — not perfectly, but closely enough that traits persist.
Darwin did not know about DNA. He only needed: like tends to produce like.
Variation
No copy is perfect.
Within any population, individuals differ — in size, colour, behaviour, shape.
Some of this variation is heritable: it gets passed on, imperfections and all.
Without variation, there is nothing for selection to act upon.
Selection
Not every variant survives. Not every survivor reproduces equally.
Resources are limited; environments are demanding.
Whatever traits make an individual slightly better at surviving and reproducing become, on average, more common in the next generation.
Nobody chooses. The environment simply… filters.
Darwin's Finches
In 1835, Darwin collected small birds across the Galápagos Islands — a few hundred kilometres off the coast of Ecuador.
Back in London, ornithologist John Gould identified them: not varieties of one species, but more than a dozen distinct species of finch — found nowhere else on Earth.
Each island's finches had a differently shaped beak: thick and crushing, thin and probing, long and curved.
One Ancestor, Many Beaks
The beaks matched the food available on each island: hard seeds, soft fruit, insects, nectar.
Darwin's explanation: all the finches descended from a single ancestral species that reached the islands long ago.
On each island: beak-shape variation + a different food source + generations of selection → a different finch.
One ancestor. One mechanism. Many birds.
Why Finches Matter
The finches are not proof of evolution by themselves — but they are an unusually clean illustration of the engine in action, isolated on real islands, at a scale a visitor could actually observe.
Replication (finches breed finches), variation (beak shapes differ), selection (local food decides who thrives) — nothing more was needed.
So here is a strange question. Does this engine only run on carbon?
Universal Darwinism
Darwin described biological evolution. But notice: nothing in the three-step engine mentions DNA, cells, or carbon chemistry.
It only requires: entities that are copied, imperfectly, and then differentially filtered.
Richard Dawkins called this generalized claim Universal Darwinism: wherever you find replication + variation + selection, you will find evolution — in biology or anywhere else.
Beyond Carbon
The pattern shows up again and again, once you know to look for it:
your immune system selects among antibody variants for the ones that bind an intruder;
computer scientists breed populations of candidate solutions in “genetic algorithms”;
markets select among competing products and firms.
Could the same be true of ideas themselves?
Dawkins' Reframing
In 1976, Richard Dawkins published The Selfish Gene — and reframed the whole story.
It is not organisms that are the true unit of selection, he argued. It is the gene.
“We are survival machines — robot vehicles blindly programmed to preserve the selfish molecules known as genes.”
Replicators and Vehicles
Two roles, easy to confuse:
Replicator — the thing that gets copied. The gene.
Vehicle — the thing that is built, survives, and reproduces, carrying replicators inside it. The organism's body.
Selection acts on vehicles — but its lasting effect is on which replicators get copied into the next generation.
A Question Dawkins Asked
If the gene is just one possible kind of replicator…
…could there be others? Things that are copied, vary, and are selected — but are made of neither DNA nor protein?
Dawkins proposed there could. He needed just one more word.
The Second Replicator
“I think that a new kind of replicator has recently emerged on this very planet… I shall call it a meme.”
— Richard Dawkins, The Selfish Gene, 1976
Tunes, catch-phrases, fashions, ways of making pots or building arches — copied from mind to mind, by imitation.
Genes and Memes Compared
Gene — replicated via reproduction; vehicle: the body; variation: mutation; selection: survival & reproduction.
Meme — replicated via imitation; vehicle: brains, books, screens; variation: mis-copying, reinterpretation; selection: attention, memorability, usefulness.
Same three-step engine. A completely different substrate.
Ideas in Motion
We began with finches on volcanic islands and ended with an idea that may be living, right now, inside each of you.
If memes are real replicators, then minds are their islands — and culture is the ongoing experiment.
In the sessions ahead, we will look much closer at how exactly this second replicator moves: through imitation, through everyday life, through law, politics, and the machines we now build.
Cultural evolution
Imitation & development of human individual
Visual memes – glyphs, emojis, latrinalia, artefacts &c.
Earworms, mindworms, moving memes::Fashion, Style, Art &c.
Mediated memes, Urban memes
Memplexes
Memes in Law, Politics & Propaganda
Memes & Cyberspace
Memes, Memecoins, AIs
Memes revisited
The Congress
Medienhaus Chronicle ⁄ Kastalia Project (VK & KuM only)
The story begins ...
Formats
Artefacts
Kastalia KMS
In Josef Knechts handschriftlicher Übersetzung / Kastalia Archiv / Hermann Hesse, The Glass Bead Game
Kastalia instances
gardens.digital
fibel.digital
Grunewaldstrasse cafeteria
kastalia.medienhaus.udk-berlin.de
Narrative behind Kastalia KMS
the project of 2nd Archive resp. "2nd Foundation" germinated for many years
ultimate impetus for creation of KMS given by need of VK and KuM study programs for a unified aggregator and generator of Einführungveranstaltung (EV) presentations
first EV in winter semester 2016/2017 at url https://cloud.medienhaus.udk-berlin.de/community
first public demonstration at UdK Zukunfstag 2019
deployment of other instances in 2021
handbook generator (SQL->LaTex->PDF) deployed in Summer Semester 2022
Implementation
Perl Dancer app behind nginx proxy
PostgreSQL database with just three tables: "knots", "bounds" and "zeitgeist"
all knot-related ad hoc information stored in knots.attributes hstore
main predicates: "is_parent", "has_bookmark", "contains" common to all instances, certain predicates "is_uttered_as", "illustrates", "is_symbiont_with" etc. instance-specific)
Philosophy
Kastalia is composed of "knots" which have attributes and relations ("bounds") to other "knots".
Everything (e.g. user, institution, event, presentation slide, fairy tale, musical piece) is a knot.
A knot is a "Ding an sich".
Bounds are predicated. (hello sir Berners-Lee, hello Wikibase, hello "semantic web")
A knot is presented to an observer (e.g. human user) by means of a template.
Combination of a knot, other knots which are bound to it and a template yield "Ding für uns".
On Tracking : Dancing Piano II.
Pulse: How Fast Does a Body Move?
Welcome: Movement Has a Pulse
Every movement, however free it looks, has a pulse. A step, a swing of the arm, a beat of a wing — each repeats, and each repetition takes time.
This session gives you the physical vocabulary for talking precisely about that pulse: period, frequency, and speed — and why they are not the same thing.
By the end, you'll be able to describe your own heartbeat, your own gait, even a hummingbird's wing, in exactly the same terms — which is also the first step toward tracking movement and turning it into sound, piano included.
Period and Frequency
A period (T) is the time it takes for one complete cycle of a repeating motion — one full step, one full wingbeat, one full swing.
Frequency (f) is simply how many of those cycles happen per second: f = 1 / T. If a step takes half a second, its frequency is 2 steps per second.
Frequency is measured in Hertz (Hz) — one Hz means one complete cycle every second. A slow, heavy motion has a long period and a low frequency; a quick, light motion has a short period and a high frequency.
The Pendulum: A Clock Made of Motion
The simplest periodic body in physics is the pendulum: a weight swinging from a fixed point.
Its period depends only on its length and on gravity — not on how wide it swings, and not on how heavy the weight is. A longer pendulum always swings more slowly than a short one.
This single fact — longer means slower — turns out to govern almost every example in today's session, from a child's leg to a hummingbird's wing.
The Heartbeat: Your Own Metronome
The most intimate periodic motion you own is your own heart: roughly one beat per second at rest — about 60-80 beats per minute, close to 1 Hz.
Under exertion — dancing included — that frequency can double or more, while the body's other rhythms (breath, stride) often tangle into related, overlapping periods.
Long before any machine measured it, the heartbeat was probably humanity's first unit of tempo.
Walking: Cadence and Stride
Walking is periodic too: dancers and gait researchers alike call its frequency cadence — steps per minute. A relaxed walk sits around 100-120 steps/min, roughly 1.7-2 Hz.
But cadence alone doesn't tell you how fast someone is actually travelling. A tall person and a short person can share the same cadence and still cover very different distances per step.
Which raises the question underneath this whole session: if frequency alone isn't speed, what else is needed?
Running and Sprinting: Frequency Is Not Speed
Speed is frequency multiplied by how far each cycle carries you — its amplitude, or in gait terms, the stride length.
Sprinters increase both: a faster leg-cycle frequency and a longer stride. Elite sprinters reach cadences above 4-5 steps/second with strides over two metres — two independent quantities, frequency and amplitude, multiplying into genuinely extreme speed.
Keep this multiplication in mind: speed = frequency × amplitude. It will come back, almost unchanged, when we get to the piano.
Why Small Creatures Move Faster
Remember the pendulum: a longer pendulum swings more slowly. The same logic applies to limbs and wings, which behave like pendulums built out of muscle and bone.
A short limb (or a short wing) is a "short pendulum" — it naturally wants to oscillate fast. A long limb is a "long pendulum" — it naturally wants to oscillate slowly.
This is why small animals move in a blur and large animals move almost in slow motion — not because of effort, but because of size itself.
Extremes in the Animal Kingdom
A hummingbird's wings beat around 50 times per second (50 Hz) — fast enough to hover motionless in the air.
A housefly's wings beat even faster, well above 150-200 Hz — too fast for the human eye to resolve as individual strokes.
Contrast that with an albatross, gliding on wings that may flap barely once per second, or an elephant's unhurried gait, with a period closer to two full seconds per step.
Same physics, three orders of magnitude apart — purely a function of size.
The Piano String Analogy
A piano string is, again, a pendulum in disguise — just a vibrating one instead of a swinging one.
Short, thin, tightly-strung strings (the piano's high notes) vibrate at a high frequency. Long, thick, loosely-strung strings (the low notes, wound in copper wire) vibrate at a low frequency.
Short and light vibrates fast; long and heavy vibrates slow — the exact same rule that separated a hummingbird's wing from an elephant's leg.
Tempo in Music: BPM as Period
Music names the same quantity differently: tempo, measured in beats per minute (BPM).
A tempo of 120 BPM means one beat every 0.5 seconds — a period of half a second, a frequency of 2 Hz. A slow adagio at 60 BPM is a 1-second period, exactly the frequency of a resting heartbeat.
Tempo is simply frequency wearing a musician's name.
Entrainment: When the Body Syncs to the Beat
Entrainment is what happens when an independent periodic body — a footstep, a heartbeat, a swinging arm — locks its own frequency onto an external one, such as a musical beat.
This is most of what dancing is, physically: a body voluntarily tuning its many natural periods — stride, swing, sway — to a shared external frequency.
Which means a dancer's movement is not just inspired by music — it can be measured, cycle for cycle, against it.
Tracking the Untrackable
Measuring a heartbeat or a pendulum is easy. Measuring a dancer's full-body period, in real time, on stage, is not — which is exactly the technical problem this course is named after.
Modern tools attack it from several angles: wearable accelerometers (measuring acceleration directly), motion-capture markers, and camera-based pose estimation (tracking joints frame by frame from ordinary video).
Each gives a stream of numbers over time — a position, a velocity — from which period, frequency and amplitude can all be recovered.
From Movement to Sound
Once a movement's frequency is captured as a number, it can be mapped — frequency to pitch, amplitude to volume, period to rhythm — onto a piano, real or sampled.
This is the project underneath Dancing Piano: using the exact physical quantities introduced today — the same ones that separate a hummingbird from an elephant — to let a tracked body play an instrument with its own motion.
Every session from here builds, piece by piece, toward that pipeline: track, measure, map, sound.
Closing: What's Next
Today's vocabulary — period, frequency, amplitude, speed, entrainment — is the physical foundation for everything that follows in this course.
Next sessions will get hands-on with the tracking side: how to actually capture a body's movement, and how to turn the resulting numbers into music.
Before then: notice your own rhythms this week — your stride, your pulse, the tempo of whatever you happen to be dancing to.
Onboarding
Before your first session, there are a few quick steps to take care of.
Onboarding consists of a 0th step and two further steps.
0th Step: Install an App
Choose whichever fits your device:
Computer — install Element.
Mobile — install Element X.
Browser — no install needed, just open udk.ai/classroom.
Step 1: Log In
Log in (or register) using the app you just installed.
UDK students — specify medienhaus.udk-berlin.de as your homeserver.
Motivated guests — Next week, You can request a free udk.ai account.
Step 2: Join the Room
Once logged in, join the seminar's room.
Either search for it directly, or wait to be invited into it.
Room names:
Session Structure
Applies to both two-hour Wednesday seminars — AI-Literacy (14:00) and Memetics (16:00).
14:00 / 16:00 — doors open
14:15 / 16:15 — Begrüßung
14:30 / 16:30 — check-in ends
15:00 / 17:00 — minute of silence ***
15:30 / 17:30 — checkout starts
15:45 / 17:45 — individual requests / onboarding help
*** minute of silence divides the session into analogue (off-the-microphone) and digital (fully archived) sections
