4 Aug 2024/42 min/Hans Dehlinger Interview
M. Low X Hans Dehlinger Interview // Expanded.Art
An 84 year old pioneer who was present at the birth of computer art in 1960s Stuttgart connects six decades of generative practice to his new Fibonacci collection with Expanded.Art.
Hans Dehlinger, 84 and nearly 85 at recording, is the oldest guest in the show's run and the only one who was in the room when computer art began. As an architecture student in Stuttgart in the early 1960s, before the university even had a computing center, he studied alongside the movement's founders: Frieder Nake, Georg Nees and the philosopher Max Bense, whose advice he still follows. Do not talk about art, talk about aesthetic events, because art is a loaded thing and an aesthetic event is specific. He learned ALGOL from a professor teaching engineers, applied it to a design project, and produced 2,500 solutions where one building was expected. The gigantic failure that followed became the turning point of his career.
The middle of the conversation is living history. Computer art was dismissed for decades by artists and every art organization worldwide (this is no art, there is no artist behind), while his drawers filled with plotter drawings nobody wanted. SIGGRAPH conferences, 40,000 attendees bused three kilometers to lunch in New Orleans or Los Angeles, were the lifeline: the discovery that you were not wrong and you were not alone. MLow draws the parallel out loud to today's crypto conferences like Consensus in Austin, and Dehlinger agrees completely, predicting MLow will keep those friendships for life the way he has.
The occasion is FIBO, his new collection with Expanded.Art, which MLow collected four or five of and left running on his screen for a day. It is Dehlinger's first project in p5.js, and its Fibonacci premise reaches back to Bense's old question of whether beauty can be calculated, answered by Birkhoff with order over complexity. His hypothesis, which he is careful to call unprovable, is that transforming the hidden order of mathematics into artistic form lets the output benefit from that order. Along the way he recalls Harold Cohen snapping write your own at a student who asked to download his code, warns that AI's speed will be underestimated exactly as computing's was, and closes with an orange tree in Jean-Pierre Hebert's garden and a lesson in humility.
What this episode teaches
- Talk about aesthetic events, not art
Dehlinger still follows Max Bense's advice from the 1960s: art is a loaded word, so discuss the aesthetic event instead, a specific and assessable thing. It remains useful vocabulary for anyone defending generative or AI work.
- The rejection can be the turning point
Offering 2,500 combinatorial solutions where architecture demanded one got his project dismissed outright, and he names that failure as the pivot of his entire career. A method that is currently unwelcome can still define fifty years of practice.
- The AI argument is a sixty year rerun
Artists and every art organization worldwide once said of computer art: this is no art, there is no artist behind. Nobody questions it today. The one guest who lived the first round expects the current AI fight to end the same way.
- We underestimate the speed, every time
Everyone involved with computers from the early days underestimated how fast the field would develop, and Dehlinger says plainly that he is afraid the same thing will happen again with AI.
- Conferences cure isolation, not just ignorance
The value of SIGGRAPH was not the talks. It was suddenly finding out that you were not wrong and you were not alone, after being isolated at home, and the people met there became lifelong friends.
- Constraint is an editor
When a single plot takes two hours to two days, you are forced to boil the system down to one piece, which he says comes very close to the architect's task. Cheap infinite rendering removed a discipline worth recreating deliberately.
- Order over complexity, held as hypothesis
From Bense's question of whether beauty can be calculated, through Birkhoff's order over complexity: transform the hidden order of mathematics into artistic form and the output may benefit. He insists it cannot be proven, only tried.
- Write your own
He recalls Harold Cohen growing angry when a young listener asked where to download his code: write your own. Dehlinger endorses the point, and MLow connects it to the CrowdStrike style outage that had just blue screened computers worldwide through reliance on one centralized party.
The conversation, topic by topic
Present at the birth of computer art
Dehlinger, turning 85, studied architecture in Stuttgart in the early 1960s before the university had a computing center. Only much later did he realize he had been at a worldwide hotspot of new things happening, the beginning of computer art, alongside Frieder Nake, Georg Nees and the philosopher Max Bense. At the time, he says, we were very naive and just followed what we were told to do.
Aesthetic events and the architect's unease
He still follows Bense's advice to talk about aesthetic events instead of art, and admits he was always uneasy being called an artist because architecture is the mother of arts. What he does with computers, he says, is the approach of an architect toward a design problem: designing has problems, methods and goals no matter what you design.
The 2,500 solutions failure
He learned ALGOL out of pure interest from a professor teaching engineers, then applied it to a design project and pinned up 2,500 combinatorial solutions with a model of a random selection. The solutions were lines of numbers mapping to words describing a small building, a cafe in a cliff hanging on a rope, entered through the rock. It was rejected because nobody evaluates two and a half thousand answers, and he calls that rejection the turning point that sent him deeper into computing and on to UC Berkeley.
Decades of dismissal
Computers doing art were disliked by artists and every art organization worldwide: this is no art, no way, there is no artist behind. His drawers slowly filled with plotter drawings he found interesting and nobody else wanted. Asked what kept him going, he says pure interest and beauty: the universe of machine drawn lines differs from hand drawn lines, and exploring it can take a lifetime without reaching an end.
SIGGRAPH and the 40,000
As an ACM student member he attended SIGGRAPH conferences held only in New Orleans or Los Angeles, the two centers big enough for 40,000 attendees, with a three kilometer bus line ferrying people to lunch. Funding required an accepted contribution, so getting there was itself a challenge. MLow says today's crypto conferences, like Consensus in Austin weeks earlier, feel the same, and Dehlinger agrees totally, predicting MLow will keep those friendships for life.
The plotter and HPGL
He still runs an old pen plotter, and the challenge is keeping the whole stack alive: an old computer, old software and a long obsolete data format. He praises HPGL as a very clever invention, pen up, pen down, move, three commands that can do anything you can imagine. Because one plot took two hours to two days, he was forced to resolve everything down to a single piece. He only ever used the black line; colored pens never met his standard.
FIBO and calculating beauty
The Expanded.Art collection is his first p5.js work, made with help, and it runs indefinitely; MLow collected four or five and left one on his screen for a day. The Fibonacci idea goes back to Bense asking whether beauty can be calculated, and Birkhoff's answer of order over complexity. Dehlinger's careful hypothesis: transform the hidden order of mathematics into artistic form and the output may benefit from it. On why the pieces evoke landscapes, he points to the layered structure of a distant view.
NFTs through fresh eyes
He finds the NFT format highly interesting: the work is not too expensive, and the mint window closes, after which it cannot be acquired. His read on collector behavior is that people buy three, four or five so they can trade and speculate, but not too much, because the money paid is not excessive.
Harold Cohen and write your own
MLow mentions seeing Cohen's plotters redrawing his old AARON code at the Whitney. Dehlinger heard Cohen lecture in Michigan and recalls a young listener asking where to download his code; Cohen got angry and said write your own. They connect it to the Microsoft outage the previous week, where reliance on one centralized party blue screened machines worldwide, and Dehlinger notes darkly that the rockets falling on Ukraine have software behind them too.
Old languages and always learning
He programmed a CDC 6400 in assembler, shifting 64 bits around, and made drawings in SNOBOL4 because HPGL is a string. A conference audience burst out laughing at the crummy language, and he silenced them by asking who could still name a language they had used 20 years earlier. His conclusion: you must learn, learn, learn, and the moment you quit learning you stick with the old stuff.
AI as the challenge ahead
He has not used AI yet but calls it the challenge ahead and is completely thrilled by what he sees others doing; he wants to experience the surprises himself. MLow, an AI artist for years, offers to help him get started, and Dehlinger says he may come back to the offer, though he is temporarily handicapped after falling off a ladder: I neglected gravity, and this is something you should never try.
The Algorists and the orange tree
He recounts founding the Algorists with Jean-Pierre Hebert and Roman Verostko, both recently deceased; Hebert wrote an algorithm defining what an algorithm is. Their work hangs at the Kavli Institute in Santa Barbara, a physics institute with an artist and poet in residence, which he says better defines what a university should be. His favorite memory is an orange tree in Hebert's garden: if nature can produce such beauty in one year, then we should try to be humble.
In their words
“I want to talk about art, talk about aesthetic events instead.”
Hans Dehlinger“I was daring to offer two and a half thousand solutions instead of one.”
Hans Dehlinger“They said, this is no art. No way. There is no artist behind.”
Hans Dehlinger“I am afraid that the same thing will happen again with AI. We probably underestimate the speed of which things will go ahead.”
Hans Dehlinger“You suddenly find out that you were not wrong and you were not alone.”
Hans Dehlinger“If nature is capable in one year of producing such a beauty, then we should try to be humble.”
Hans DehlingerCite this conversation
Hans Dehlinger Interview in conversation with MLow, "M. Low X Hans Dehlinger Interview // Expanded.Art," The MLow Show, 4 Aug 2024. https://show.mlow.xyz/e/2024-08-04-m-low-x-hans-dehlinger-interview-expanded-art.html. Quoted from a machine transcript.
Quotations come from machine transcripts. Check wording against the audio before republishing.