Quacking Like A Duck

“Each reader reads only what is already inside himself. A book is only a sort of optical instrument which the writer offers to let the reader discover in himself what he would not have found without the aid of the book” (Marvin Minsky, Society of Mind, 1974).

Once upon a time in the early 18th century, one can imagine an introverted, creative young man named Jacques going about his childhood in Grenoble in the French Alps, one of nine cities experiencing rapid urbanization. Representative of the Third Estate though hardly a peasant-farmer, his father had joined a guild and made gloves in his shop, an artisan riding a wave of enthusiasm among the nobility for beautifully colored and ornate gloves.

Jacques must have gone with his father to buy the hides of young goats, kidskin, a soft, flexible, sturdy material. He surely learned that these hides were treated and turned into leather, sorted for thickness, sent for coloring of various tints. He must have watched or used himself scissors to slice finished hide into hand and fingers, a tedious and imperfect process, which a century later became automated:

“In 1834, Jouvin prepared a table classifying 320 different kinds of hands, based on measurements of wrists, fingers, and so on. Soon afterwards, Jouvin invented the main de fer, or iron hand, a mold that could cut three or four sheets of leather into glove shapes with a single motion.” (SUKHADA TATKE January 21, 2020)

Of course, Jacques knew nothing of the iron hand, but the idea of something like an iron hand may have occupied his thinking. He must have always been thinking. As an adult, Jacques found that wealthy patrons subsidized those with outsized talent of the sort he possessed.

*

Helplessly standoffish yet lonely as a boy, possibly, Jacques turned inward and fed his fascination by shape-shifting meanings in playful joy with phantoms and figments, words, images, ideas, figures flitting around like fireflies. In the outer world he stood among but not of them, as a poet of the future was to write in a dark mood.

Jacques’ father was affluent, intelligent, good-natured, aware of and grateful for his good fortune. He had established quite a lucrative business, and he must have been determined to bring Jacques out of his shell, fulfilling his obvious potential. Father in fact hired some help.

A middle-aged man was likely secured to tutor Jacques, a tutor who probably believed that happiness sprung from virtue, virtue from knowledge. Classically educated and in demand as happened during the crowning of elite tutoring, his tutor worked alone with the boy on three mornings each week, the only slots left in his schedule, perhaps stretching Jacques’ family budget.

Jacques loved learning—and drawing, likely sketching unlikely objects like ducks, flutes, maybe tools used to make gloves. Once he was rumored to have taken apart a clock and reassembled it, sketching each part in an inventory.

Jacques learned to make gloves from his father, he may have even devised improvements in methods and tools, but what he really, truly wished for was to be a clockmaker, to make cuckoos cuckoo, chimes chime, dancing maidens pirouette at the witching hour.

*

The wheel of fortune rolled on, Jacques grew into a strapping young man, and he chanced one day to meet a famous and respected surgeon, Le Cat, an expert judge of intellect who recognized the rarity of this young man’s gift. Le Cat was well known in Grenoble for his expertise in human anatomy and physiology. Some said he could diagnose the causes of pains and illnesses through palpitations and manipulations. He could read skin and bones, people said, though not everyone believed it.

Like Jacques, Le Cat felt alienation, finding his own refuge in skeletons and cadavers. When Le Cat learned of Jacques’ dreams to create lifelike animal and human figurines with preprogrammed arms and legs, the famous surgeon was delighted.

As it happened, Le Cat had an enduring interest in replicating anatomical forms and structures to use for replacement parts; he saw the world from the perspective of homo erectus, he realized, but somebody had to do it. How do we keep these human machines going? Le Cat might have imagined a future when medical innovators might make artificial hearts and lungs.

Le Cat decided to mentor Jacques in a way no professional tutor could in three mornings a week, to teach him how to be virtuous by conquering the tedium of knowledge building, to teach him everything he knew about human movement, anatomy, and physiology. From these experiences Jacques could learned to create mechanical structures to mimic living creatures—automata, robots, artificial creatures for which he would earn an enduring place in history and bring Le Cat with him.

Jacques de Vaucanson never became a clockmaker. He did, however, make a timely mark. Some say he was a crucial link in the chain that led to the Industrial Revolution, a spirit of the ilk of Henry Ford. He graduated from producing life-like toys, essentially, objects of entertainment, to inventions that would change the world.

An inventor and artist of some repute during his lifetime, Jacques piece de resistance was the invention of the metal lathe, the mother of all metal tools in the hands of skilled machinists who make gouges, chisels, knurlers, screws, tools to make iron hands, the combine of glove makers, all leading to this tool we call AI. Jacques could perhaps have had an inkling that big things were coming for his lathe:

“Between the late 19th and mid-20th centuries, individual electric motors at each lathe replaced line shafting as the power source. Beginning in the 1950s, servomechanisms were applied to the control of lathes and other machine tools via numerical control, which often was coupled with computers to yield computerized numerical control (CNC). Today manually controlled and CNC lathes coexist in the manufacturing industries.”

*

The year 1737 was productive for Vaucanson. That year he designed and created The Flute Player, a life-size figure of a shepherd who played tabor and pipe and had mastered a repertoire of twelve songs. Its original fingers were insensitive and couldn’t play the flute with human finesse, so Vaucanson gloved his shepherd’s fingers in skin. Later that year, he created The Tambourine Player and The Digesting Duck, the latter considered his masterpiece.

With over 400 moving parts in each wing, the duck’s wings could flap. It could drink water, eat, and to the delight of onlookers, defecate. Vaucanson's duck appeared to digest food, but it was a trick—it actually contained a hidden compartment of digested food. What the duck expelled was not what it ate, a mixture of water and seed. It excreted bread crumbs and green dye. Vaucanson invented the world's first flexible rubber tube while building the duck's intestines.

Jacques must have observed the delight his wealthy patrons displayed, the bursts of laughter, with a measure of ambivalence. What sort of intelligent species would gather to watch a simulated duck artificially defecate?

*****

- robot: Hello, How can I assist you?

*****

Charles Babbage worked as a professor of mathematics at the University of Cork in Ireland during the mid-19th century. With a penchant for applying mathematics to solve practical problems, in his 20s he grew irritated with inaccuracies in mathematical tables, e.g., actuarial tables, the logarithm table, etc. These tables were calculated by humans and riddled with mistakes. Babbage set to work on a calculating machine to mimic the mental actions of human tablers more efficiently and, more importantly, with greater accuracy.

The creation of automata and the invention of machines that imitated human-like behavior could not have escaped Babbage’s awareness. He may have seen or read about the quacking defecating duck. The conceptual foundation for Babbage's ideas about intelligent machines has been traced to the ancient world.

Babbage was well-aware of advances in mathematics and logic, such as George Boole's work on symbolic logic, which laid the groundwork for computational thinking. These various influences contributed to Babbage's vision of programmable machines capable of performing tasks that simulate human intelligence.

*

Alan Turing, a British mathematician who died in 1954, a victim of cruel limitations in human intelligence, earned a Ph.D. in mathematics at Princeton in 1938, just a few scant years before John McCarthy would earn his doctorate there in 1951 to be followed by Marvin Minsky in 1956. Turing is a fascinating figure in the story of AI.

Turing’s mechanical brain was a pattern-seeking tool used during WWII to crack codes generated by the Nazi’s Enigma machine, codes carrying secret information transmitted to German soldiers, sailors, and pilots, using radio waves.

Turing built the simplest computing model he could, simple but powerful enough to calculate all possible functions that can be calculated. The machine first had to discern the pattern in the incoming radio information. Thereafter, the pattern could be used to decode the message. Turing contributed his computer to increase available intelligence as a way to stymie raw power turned against humanity. The future was quickly arriving.

*****

Marvin Minsky earned a Ph.D. in mathematics from Princeton. In 1958 he was hired at MIT where he co-founded MIT’s AI laboratory with John McCarthy, the professor who originated the phrase “artificial intelligence.” Minsky worked at MIT until his death in 2016.

In 1955 Marvin Minsky and John McCarthy, the co-founder of the MIT AI lab, wrote a proposal for a summer research project at Dartmouth College. Minsky and his colleagues had a wild idea, a hunch, that general human intelligence could be simulated by a computer. The interest was in creative and competent thinking, not brilliance. Here’s what they wrote:

“We propose that a 2 month, 10 man [sic] study of artificial intelligence be carried out during the summer of 1956 at Dartmouth College in Hanover, New Hampshire. The study is to proceed on the basis of the conjecture that every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it.”

Minsky et al. had intuited the central problem with the brute-force computational paradigm dominating computer science in the 1950s and beyond that sought to create a mechanical brain to solve one, and only one, challenge (e.g., chess). In their proposal they wrote “…[with]… a well-defined problem (one for which it is possible to test mechanically whether or not a proposed answer is a valid answer) one way of solving it is to try all possible answers in order. This method is inefficient.” In their era it might have taken a computer as big as a barn to balance a checkbook.

It wasn’t just inefficiency. Task specialization was the Achille’s heel of the then dominant goal of computer science. IBM may have been designing a machine brain that would eventually play chess like a grandmaster, a one-trick pony. But human intelligence can play chess, then play piano, then design a bridge, heal the sick, and then have breakfast. Minsky was an admirer of Einstein. He quoted him at the start of his seminal book Society of Mind:

In the Dartmouth proposal in 1955 Minsky and McCarthy asserted a design principle that pointed the way to natural language algorithms. “A fairly attractive and yet clearly incomplete conjecture,” they wrote as a preface to this principle, which would take 70 years to become a reality, “is that the difference between creative thinking and unimaginative competent thinking lies in the injection of some randomness.” Ultra competent simulated thinking of the AI chess master was unimaginative and precise for Minsky and McCarthy and therefore uninteresting. Human thinking, however, was creative and imaginative, the trait that really made it human.

“The randomness must be guided by intuition to be efficient. In other words, the educated guess or the hunch includes controlled randomness in otherwise orderly thinking.”

*

Minsky, along with his colleague Seymour Papert, developed the concept of "frames," which represented a structured way of organizing knowledge that would pave the way for AI knowledge graphs and maps. Fortunately, three fields of research were converging to produce conceptual and theoretical knowledge about language, humans, and machines: grammatical transformations (Chomsky), schema theory (Pearson and Anderson), and artificial intelligence (Minsky, McCarthy) which would eventually materialize in the bot as an assistant reader and writer.

John McCarthy was completing his doctoral degree in mathematics at Princeton when Marvin Minsky began his graduate work there. McCarthy’s interest in machine intelligence was sparked at a conference in 1948 where researchers from mathematics, computation, and psychology gave talks about the possibility of machines that could simulate human intelligence. Before he met Minsky in 1951, McCarthy was interested in mechanical intelligence and worked with what he called “brain automata.”

McCarthy soon learned of Minsky’s interest in machine intelligence, and they began to engage with one another. In 1953 McCarthy accepted a position as an assistant professor of mathematics at Stanford. For the next few years, according to McCarthy, he devoted himself to partial differential equations and their analysis. In 1955, however, his path crossed with Minsky during a summer workshop at IBM.

McCarthy, often referred to as the "father of AI," was sure that the brute-force paradigm was a dead end. Instead of picking an intellectually daunting task and finding a way to design and program a mechanical brain to solve it, he was one of the first to explore commonsense reasoning in his research. Instead of making a machine to master chess, he was interested in making a machine that had enough sense to understand how to bake a cake or have a talk about a bicycle. He developed the logic-based framework called "circumscription," writing around, which aimed to represent a circumstance and a task with incomplete information. This sort of task and circumstance required human intelligence—or its simulation.

***

That an AI bot disrupting the lives of teachers in 2023 has a defecating duck in its ancestry is ironic but pungent. As Jacques must have pondered when the nobility marveled at the quacking, flapping, and defecating of his artificial duck, what is this human tendency to believe in technological fantasies, to shape-shift tools into living beings? How did Hal become a personality, a believable main antagonist in a reputable movie?

The earliest iterations of robots were simulations of how the brain controls bodily functions like digestion or manages fingers to play a flute. Jacque’s motive for gaining the knowledge needed for the shepherd and the duck was an urge to make more knowledge, knowledge Le Cat would use to advance the cause of artificial spare parts for humans. Babbage’s motive was noble—accuracy: Mechanical intelligence in algorithmic mathematical computing is simply better done by a machine. Turing worked to rescue the world from Adolph Hitler.

Over three centuries AI simulations took on an agenda far beyond the wildest dreams of the early innovators. One item on the agenda was making a mechanical brain that could compete in chess at a world class level. A second item was to create such a brain that could decode military communications among the Nazis to fight against tyranny and genocide. A third item was to make a device with common sense that could simulate the ways humans use language well enough to carry on a robust conversation.

In the 1960s and 70s AI thought leaders who chose the path of commonsense in circumstances posing ill-defined tasks as opposed to leaders who chose the path of brute-force chess master thinking in predefined and rule-governed situations worked at it until the brute-force Deep Blue breakthrough in the 1990s. AI could beat human chess masters. Then came the common sense breakthrough in 2019 when AI finally matched humans in the Winograd Challenge, disambiguating pronouns when competing against ordinary humans.

And now we have a bot we can talk to about its ancestor, the defecating duck.

Once upon a time in a future not too far away…