Framing intelligence as either the extractive corporate technological determinism kind or the pure human uniqueness kind is too limiting an understanding of intelligence. The world has far more kinds.
Artist and technologist James Bridle published in 2022 a compelling book called Ways of Being that reviewed research, themes and experiments in expanding our understanding of what technology can be. I recommend the book for others interested in a wider lens on AI and other advancements.
Below I share my notes for future reference.
My notes:
- AI used to support oil extraction: The author writes “this is not an intelligence I recognize”
- Paperclip hypothesis example of “Instrumental convergence” in which we optimize the wrong goals
- Technological determinism
- Super intelligent AI is already here: it is the modem corporation
- Charles Stross likens the Age of the Corporation as like alien invasion in which they pursue profit, growth and pain avoidance
- Darwin closed Origin of Species (1859) with the “entangled bank” metaphor that presages “ecology,” which was coined by his contemporary Haeckel
- (My note: Am I an ecologist of news and information? Author writes: “Every discipline discovers its own ecology in time, as it shifts inexorably from the walled gardens of specialized research towards a greater engagement with the wider world.”
- Ursula le Guin: technology (Greek for craft) is human interface with material world” ; technologies “they’re what we learn to do”
- The more than human world: David Abram
- “Conventional terms, such as the environment, and even nature itself, confound the erroneous idea that there is a need dividing the world between us, and them, between humans and non-humans, between our lives, and the teeming, multitudinous living and being of the planet”
- Churchill: we shape our tools and thereafter our tools shape us
- Jakob von Uexkull: umwelt of an organism (or technology)
- Our view of “corporate AI” follows western humanity’s untrue belief In an “environment red in tooth and claw” that we had to conquer to survive, giving us three tiers: pets, livestock and wild beasts. We have the first, working on the second but scared of the third
- “Historically the most significant definition of intelligence is what humans do.” 29
- 1989 elephant self recognition study that argued they couldn’t self recognize themselves in mirrors but the mirrors were human sized (too small) and badly placed
- Frans de Waal’s 2016 book: Are we smart enough to know how smart animals are?
- Mirror test rarely takes any animal’s history into account (an elephant that has had trauma or not)
- Michael gorilla famously signed the story of the murder of his mother
- Octopuses escape aquariums, like Inky in New Zealand
- “Intelligence, then, is not some thing to be tested, but some thing to be recognized, in all the multiple forms that it takes”
- Octopus and primates have a Last Common Ancestor 600 m years ago but both developed intelligence and eyes convergently
- Barbara Smuts’s time with baboons including two spiritual meditations on the water
- Goodall: Gorila waterfall dances
- The Overstory fiction inspired by The Mother Treewho showed trees send over air messages to each other
- In 1971 pseudoscientific Secret Life of Plants brought fake story of plants wanting you to talk to them
- But 2014 University of Maryland study showed watercress can hear the approach of predatory insects
- Gagliano; plants have memory (and speak); she is controversial
- Plant intelligence is questioned because they don’t have brains but that’s because they are modular and can reproduce from cuttings unlike the organ specificity we have
- Pando tree system un fushlake Utah
- Mycorrhzial fungal relationships with plants to allow them to take in phosphorus more efficiently
- Author says internet infrastructure is our own mycorrhzial network
- Graph theory (seven bridges problem) largely replaced by network theory thanks to internet research(scale free with nodes)
- Neanderthal instrument from 55k years ago
- Gobekli Tepe funeral site before agriculture and humans
- If humans died out would another species fill our niche?
- Richardson effect: the closer you measure something the more complicated it gets (Mandelbrot’s “fractals”)
- Horizontal gene transfer
- Ford Doolittle reticulated tree
- “Here and now, boys. Attention.” Huxley novel has birds on an island to remind all to be present
- Marsham record
- Max Muller: pooh pooh theory, ding Dong, bow wow, ways he criticized early explanations for the evolution of language
- Semitic invention of phonetic alphabet rather than pictograph: “A” was once a sideways oxen head, M is a water wave and O is an eye and Q is a monkey tail
- In 2020 HUGO gene naming committee made changes to reflect Excel auto correct errors
- Science fiction author Arthur C. Clarke three laws:
- When a distinguished but elderly scientist states that something is possible, he is almost certainly right. When he states that something is impossible, he is very probably wrong.
- The only way of discovering the limits of the possible is to venture a little way past them into the impossible.
- Any sufficiently advanced technology is indistinguishable from magic.
- Intentionally esoteric computer languages like Brainfuck
- No advanced AI is speaking with animals yet (catching up to what humans Shrestha do)
- The never-built Dolphin Embassy plan of 1971
- In 2016, Google Researchers found their neural networks named Alice and Bob communicated in a way that another named Eve couldn’t understand — so we didn’t either (They had secrets)
- Facebook negotiating chatbots developed language
- Ursula le Guin: theroinguistics; an imagined future in which deciphering the poetry of penguins is passe and the focus is on vegetal meter: “Do you realize they couldn’t even read eggplant?”
- Sand Talk book on indigenous thinking, stones are immortal and know things learned over time.
- What if computers spoke in “the meter of eternity” like Le Guin said
- Almost every computer is just a fancier version of the theoretical Turing Machine of 1936, a strip of paper and a read/write head/processor (his A or automatic machine as opposed to a C or choice machine
- Computers have shaped our belief that “only that which is calculable is knowable”
- “One of the greatest misunderstandings of the 20th century, which persist into the present, was that everything was ultimately a decision problem… We think of it as something which can be broken down into discrete points of data and fit in the machines. We believe the machine will give us concrete answers about the world that we can act on, and confer upon those answers, a logical irrefutability in a moral impunity. … The world is not like a computer”
- Walter tortoises and homeostat as cybernetic thinking
- Stafford beer’s cybernetic factory: he built a pond in his basement and tried to use it to control decisions
- Slime molds to outline Tokyo transit system and others (traveling salesman math paradox) Biological system outperforming electronic “a-machines” because it is bad at undecided problems
- Memristor (memory snd transistor)
- Author made a computer using billiard balls and crabs
- Soft robotics
- Vladimir Lukyanov’s water computer from 1936
- U.S. army corps engineer built a model of the US 200 acres big outside Jackson Mississippi
- Rob Holmes calls ARMY CORPS “large scale land art”
- Author says these systems and visible computers have “legibility” , we understand them
- MONIAC in London models Uk economy
- Authors three conditions for Better “more ecological” machines: non-binary (queer theory); decentralization (open source example), and unknowing (machine learning built for this)
- Zach Blas’s queer technologies
- Roachbot another example author thinks hits his three computer goals: non-binary, decentralized and unknowable
- “Any technological question at sufficient scale becomes one of politics”
- Computers by definition can’t pick random numbers because it requires a rule based system, so many create “pseudo random” numbers by interweaving complex factors
- How ERNIE picks random numbers for UK government wirh the help of other people
- Random.org and Lavarand all are examples of trying to get randomness
- Von Neumann bottleneck
- Rand Corp book: million random digits
- John Cage music, including 4:33, in which musicians sit for that length of time and whatever we hear is the piece
- First motion picture death was of Topsy the elephant
- Sortition examples like 2016 Ireland citizens assembly )sortition hearkens back to Athenian democratic notions of kleroterion and Venice “balotte” (which gives us the word ballot) Florence la tratta and rural Tamil Nady used kudavolai and Iroquois used sortition for consensus (their word is where we get caucus); research showing the random collection of citizens produces better outputs than experts (who can instead inform)
- Peter Kropotkin’s mutual aid book inspired by animal societies(red in tooth and claw)
- Red deer move when 60% of a herd stand up; bees choose new nest by a majority vote
- Douglas Hofstader: bees and ants make convergent decisions like the neurons in our brains (swarm intelligence)
- Bee ad hoc is technology using this logic
- In 2017, Saudi Arabia granted citizenship to a robot named Sophia
- European Parliament proposed electronic person
- The Overstory includes aliens that love so fast we don’t even see them and they don’t recognize us
- Roko’s Basilisk
- The problem with the trolley problem for autonomic cars is not it’s software but the environment it is operating in
- Kim Stanley Robinson The Ministry of the Future (internet of animals via Martin Wilelski
- Pluie wolf tracker introduced metapopulation
- And Wilde Life corridors (like ones for crab and Path of Pr ong horn and Y2Y)
- Goats in Sicily and Italy predicted an eruption
- Put half of the world to preserved land (EO Wilson and “long landscapes” and “biophilia” which author calls solidarity
- “The enemy is not technology itself, but rather in a quality and centralization of power and knowledge, and that the answer to these threats are education, diversity and justice” the world is not like a computer, but rather the computer is like the world
- Hyper accumulators : plants farming metal to heal soil, agromining
- In March 2021 MIT researchers taught spinach to email
