AI4Every1 : Chapter 1 (Part 1)

Historical Overview & Key Milestones of A.I.

"Computers will understand sarcasm before Americans do."

-- Geoffrey Hinton
(Nobel in Physics, 2024)

Alright, before we dive headfirst into the silicon brains and neural networks of today, let's hop into our metaphorical time machine (sadly, not AI-powered... yet, give the engineers a minute!). Every great story has a beginning, and the dream of artificial intelligence isn't some flash-in-the-pan idea born in a Silicon Valley garage last Tuesday. Nope, humans have been fantasizing about intelligent machines, automatons, and generally bossing around non-biological entities for centuries, maybe even millennia! It's a tale woven through myths, legends, philosophical debates, and some surprisingly clever ancient engineering. So, grab your historian hats (they look smashing on you, by the way) as we excavate the very foundations of AI thought, long before computers even existed.

Early Conceptualizations of A.I. (Or, When Myths Had Robots)

Long before the first transistor hummed or the first line of code blinked on a screen, the idea of artificial life was simmering away in the human imagination across diverse cultures. It seems we've always been fascinated (and perhaps a tad intimidated) by the possibility of creating intelligence that wasn't, well, born.

Think ancient Greece, for instance. They weren't just busy inventing democracy, philosophy, and drama; they also dreamt up Talos [1]. Not some friendly neighborhood uncle, but a giant automaton forged from bronze by the god Hephaestus himself, programmed (in a manner of speaking) to patrol the shores of Crete and greet unwanted guests with massive boulders. Talk about an early, rather aggressive, border security system! And let's not forget Hephaestus's other side projects -- golden mechanical handmaidens who could anticipate his needs [2]. Early smart assistants, perhaps, minus the annoying software updates?

Then there's the rich tapestry of Jewish folklore, which gifts us the Golem [3]. The most famous iteration is the Golem of Prague, supposedly sculpted from river clay by Rabbi Loew back in the 16th century to act as a protector for the Jewish community. Animated by mystical means (insert 'divine algorithm' joke here), the Golem was immensely powerful but notoriously difficult to control -- serving as an early, rather muddy, cautionary tale about the potential dangers of powerful artificial creations going off-script. (Disclaimer: While Rabbi Loew is central to the legend, attributing the physical sculpting of the Golem to him is part of the folklore; actual historical sculpting details are less certain! [4])

Peacock-shaped Hand Washing Device: Illustration from The Book of Knowledge of Ingenious Mechanical Devices (Automata) of Inb al-Razza al-Jazari (recto), 1315. Syria, Damascus, Mamluk Period, 14th Century. Opaque watercolor and gold on paper. (Source)

Meanwhile, over in ancient China, remarkable stories circulated about mechanical ingenuity that would make modern roboticists jealous. Consider the tale of Yan Shi, an 'artificer' who, according to texts dating back centuries BCE (specifically the Liezi), presented King Mu of Zhou with a life-sized, startlingly realistic automaton [5]. This mechanical marvel could apparently sing, dance, and even flirt (reportedly causing some royal consternation). This highlights a long-standing fascination not just with automated function, but with replicating the very appearance and social behaviors of life.

And while the term 'AI' is recent, let's not overlook the profound intellectual heritage of India. While direct 'robot' myths might be less prominent than in Greek tales, ancient Indian philosophy delved deeply into the nature of consciousness, logic, and intelligence itself. Think of the astonishingly sophisticated rules of Sanskrit grammar formalized by Pāṇini around the 5th century BCE [6]-- essentially an incredibly complex, rule-based system for language generation and analysis, a precursor to computational linguistics! Furthermore, epics and historical accounts sometimes mention intricate mechanical devices and automata (like the legendary flying vimanas or automated figures in royal courts) [7], suggesting a long-standing cultural engagement with advanced technology and the concept of created, functional beings.

Fast forward to the Islamic Golden Age (roughly 8th to 14th centuries CE), where inventors like the brilliant Al-Jazari [8] weren't just theorizing -- they were building intricate automata! His 12th-century "Book of Knowledge of Ingenious Mechanical Devices" describes programmable musical fountains, automated peacock statues, and even a famous hand-washing automaton which presented soap and a towel sequentially. These weren't 'thinking' machines, but they were groundbreaking examples of automated systems using principles like camshafts and sequences -- crucial steps on the long, winding road towards modern robotics and AI.

So you see, the ambition to breathe life into the artificial, to create helpers, companions, entertainers, or protectors from inanimate materials, is practically woven into our collective human DNA (or perhaps our collective source code?). These ancient dreams, myths, and early mechanical wonders laid the conceptual groundwork, planting the seeds for the computational revolution that would sprout centuries later. $\gamma$

References

1. Stanford Report, Wiki, The Theoi Project, Greek Mythology
2. The Collector, Kosmos Society, Theoi
3. Wiki, Britannica, Jewish Virtual Lib, EBSCO,
4. GargWiki, Jewish Book Council, Jewish Studies, My Jewish Learning, Humanities Mag, YIVO
5. Wikiwand, Wiki, Ancient Origins, Kevin LaGrandeur, Brewminate
6. Wiki, Internet Archive, Science India Mag, ResearchGate
7. Book by D. H. Childress, Scribd, Wiki (Vimana, Automaton), Wisdom Lib (1, 2)
8. Muslim Aid, MH (1, 2), MIT Press, OE, CT, HoI, BAF, MFAB, Book, Wiki, AO

The Calculating Engines: Paving the Way for Brains

Okay, so humanity dreamt of smart machines, but dreams alone don't compile code or crunch numbers. To get to anything resembling AI, we first needed machines that could actually calculate stuff -- reliably, and preferably faster than a commerce student during an exam trying to balance a sheet. The journey from simply counting pebbles to complex computation is quite the story, essential groundwork for the 'thinking machines' to come.

It arguably started thousands of years ago with humble tools like the Abacus [1]. Don't underestimate this beaded frame! It was the original calculator, a staple across Asia and Europe for centuries, proving a fundamental concept: we could build physical aids for complex mental tasks. Think of it as the ancient world's spreadsheet software, just requiring more finger dexterity and less electricity.

Let's zoom forward past slide rules and other ingenious contraptions to the 17th century. Enter Blaise Pascal, a French genius who, around 1642, built the Pascaline [2] to ease his father's tax calculation woes (proof that bureaucracy, like necessity, is the mother of invention!). It was a mechanical calculator with gears and dials. A few decades later, Gottfried Wilhelm Leibniz, another polymath powerhouse, created the Step Reckoner [3], which could handle multiplication and division. These were like intricate clockwork music boxes, but for arithmetic -- clever, but limited.

A Pascaline signed by Pascal in 1652. By Rama, CC BY-SA 3.0 fr, WikiMedia

The real paradigm shift, the moment calculation started dreaming of becoming computation, arrived with a brilliant, albeit famously grumpy, Englishman: Charles Babbage in the 19th century. He first designed the Difference Engine (parts of which were actually constructed) [4] to automate the tedious process of producing mathematical tables (think log tables, but more complex). But his magnum opus was the Analytical Engine [5]. Though never fully built in his lifetime (funding issues -- some things never change!), this was revolutionary. It was designed to be a general-purpose, programmable computer using punched cards for instructions! It had a 'mill' (the processor) and a 'store' (memory) -- the basic blueprint of computers today. And who truly grasped its potential? Ada Lovelace, daughter of Lord Byron and a gifted mathematician. She wrote what are considered the first algorithms intended for a machine, envisioning the Analytical Engine manipulating not just numbers but symbols, potentially creating music or art. She saw beyond calculation to computation -- a visionary leap!

The pace picked up towards the end of the 19th century with electricity entering the picture. Herman Hollerith devised an electro-mechanical tabulating machine using punched cards to process the 1890 US Census data at lightning speed (for the time) [6]. Its success led to the formation of a company that eventually morphed into IBM -- yes, that IBM. These tabulators became the workhorses of business and government data processing for decades.

But the true electronic dawn broke mid-20th century. Alan Turing's theoretical work on the Turing Machine [7] in the 1930s laid the conceptual foundation for what a universal computing machine could do. Then, driven partly by the immense computational needs of World War II (like cracking codes -- shoutout to Bletchley Park!), we saw the emergence of the first electronic computers. While Colossus helped crack codes in the UK [8], the ENIAC (Electronic Numerical Integrator and Computer), completed around 1946 in the US, is often hailed as the first general-purpose electronic digital computer. This room-sized behemoth, filled with vacuum tubes, wires, and requiring manual reprogramming, was a monumental leap [9]. It could perform calculations thousands of times faster than any human 'computer'

Lovelace's diagram from "Note G", the first published computer algorithm. By Ada Lovelace -, Public Domain

(which, back then, was a job title for people doing calculations!).

From the Abacus to ENIAC, each step wasn't just about faster math; it was about automating logic, storing instructions, and manipulating information in increasingly sophisticated ways. These calculating engines were the essential hardware -- the 'body' -- waiting for the 'mind' of artificial intelligence to be conceived. They proved that complex processes could be mechanized, setting the stage perfectly for the pioneers who would soon ask: "Can we make these machines not just calculate, but think?" $\delta$

References

1. Abacus History, Wiki, FC, CueMath
2. Wiki, Educal, LTS, Britannica, Reddit
3. Wiki, HoI, Hannover, Fiveable
4. Britannica (CB, DE, History of Computing, First Computer)
5. Britannica, Lovelace
6. NMAH, IBM, USCB, CHM
7. SEP (The Man, The Machine, The Problem, The Work)
8. Wiki, GCHQ, TNMOC, EBSCO
9. Britannica Kids, CHM, Britannica

The Official Birthday Party: AI Gets a Name (and a Test)

So, we had the ancient dreams, the philosophical musings, and finally, the calculating machines like ENIAC blinking away in university basements, humming with potential. The ingredients were simmering. Now, it was time for someone to actually stir the pot and officially declare, 'Let there be AI!' This pivotal moment, the transition from mere omputation to the pursuit of intelligence, happened remarkably quickly after those first electronic brains flickered to life [1], thanks largely to a few brilliant minds asking audacious questions.

First, let's circle back to our codebreaking genius, Alan Turing. This wasn't just any mathematician; Turing was a Cambridge logician whose theoretical 'Turing Machine' had already defined the limits of computability, and whose practical work at Bletchley Park during WWII was instrumental in cracking German codes (quite literally helping save the world -- no pressure for his next act!). In his landmark 1950 paper, "Computing Machinery and Intelligence" [2], published in the journal Mind, Turing didn't just ponder "Can machines think?"; he cleverly sidestepped the endless philosophical debate by proposing a pragmatic alternative: the Turing Test (or 'Imitation Game' as he called it). Turing considered the original question "too meaningless" and sought a replacement "expressed in relatively unambiguous words". This reframing, focusing on observable behavior rather than unobservable internal states, provided a practical starting point for a new field, avoiding the immediate need to definitively define 'thinking' itself.

Picture three separate rooms. Room 1 has the Human Interrogator. Room 2 has a Human Participant. Room 3 has the AI Machine. The Interrogator communicates with both Room 2 and Room 3 via text terminals only, perhaps labelled ‘A’ and ‘B’, without knowing (Image generated by Imagen 3 and then edited.)

Let's revisit the setup [3]: You, the astute human judge (interrogator C), chat via text with two hidden players -- one human (A), one machine (B). Your mission, should you choose to accept it, is to figure out which is which. Can you probe their knowledge of yesterday's cricket match? Ask for their opinion on the best way to make Aloo Posto? Request a sonnet about Kolkata monsoons? While these examples illustrate the conversational nature, Turing envisioned an unrestricted test covering "almost any" subject, assessing broad conversational ability rather than narrow expertise. If, after a good grilling (Turing suggested about 5 minutes), the machine can fool you into thinking it's human often enough (he ballparked that by the year 2000, an average interrogator would have no more than a 70% chance of making the correct identification after five minutes, meaning the machine fools them 30% of the time), then it passes the test. It's crucial to remember Turing wasn't claiming this proved consciousness or 'real' thinking. Rather, it was an operational benchmark: can a machine behave intelligently enough in conversation to be indistinguishable from a human? It was a brilliant, provocative idea that kicked off decades of debate, inspired generations of AI researchers (especially in natural language processing or NLP [4]), and remains a powerful cultural symbol for machine intelligence, even if its practical value as a definitive test is still argued over a cup of tea. While its direct application as a rigorous scientific benchmark has been questioned and often replaced in modern AI research, the test's cultural and philosophical impact persists, framing public discourse and ethical considerations surrounding artificial intelligence.

While Turing provided the philosophical spark and a benchmark, the actual naming ceremony and formal launch of AI as a research field happened a few years later, fueled by post-war optimism and the growing power of computers. The venue? Dartmouth College, USA. The event? The legendary 1956 Dartmouth Summer Research Project on Artificial Intelligence [5]. This wasn't just any academic conference; it was an ambitious, month-long brainstorming session intended to kickstart the entire field, marking the birth of AI as a research discipline. Let's meet the key players who convened this intellectual feast:

Name	Affiliation (Then)	Key Role / Contribution
John McCarthy	Dartmouth College	Initiated and co-organized the workshop; Coined the term "Artificial Intelligence" for the event and field; Defined AI as "the science and engineering of making intelligent machines"; Later invented Lisp programming language.[6]
Marvin Minsky	Harvard University	Co-organized the workshop; Co-founded MIT AI Lab (with McCarthy); Pioneer in AI, robotics, knowledge representation (Frames), cognitive science, early neural networks. [7]
Nathaniel Rochester	IBM Corporation	Co-organized the workshop; Chief architect of IBM 701 (first mass-produced scientific computer); Provided crucial engineering / hardware perspective; Wrote first symbolic assembler. [8]
Claude Shannon	Bell Telephone Laboratories	Co-organized the workshop; Renowned as the "father of information theory"; His work provided foundational mathematical tools for digital circuits, communication, computation, and uncertainty relevant to AI. [9]

Their proposal for the workshop radiated an almost infectious optimism (some might say hubris!), boldly stating: "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.". This confident assertion, aimed partly at securing funding from the Rockefeller Foundation, reflected not just scientific belief but also a necessary strategic posture to launch a speculative field. They aimed to make significant headway that very summer on incredibly complex topics: how to make machines use language, form abstractions and concepts, solve problems previously reserved for humans (like playing chess or proving mathematical theorems), and even achieve self-improvement and creativity!

Five of the original participants in the 1956 conference on artificial intelligence at Dartmouth gather at the 50th anniversary in 2006. From left are Trenchard More, John McCarthy, Marvin Minsky, Oliver Selfridge, and Ray Solomonoff. (Photo by Joseph Mehling; Source: Dartmouth College Website)

Did they crack general AI in those few weeks? Of course not! (If they had, this book would be much shorter, and possibly written by an AI 😉). The initial burst of optimism soon met the harsh realities of computational complexity. However, the Dartmouth Workshop was immensely significant. It officially christened the field , brought together its founding figures , established a shared (if broad) vision , and laid out the initial research agenda -- focusing on areas like search algorithms, natural language processing, machine learning, and logical reasoning. Although major technical breakthroughs were not achieved during the workshop itself, and collaboration was less integrated than initially hoped , its primary success lay in the social act of definition. By providing a name, a core group of proponents, and a foundational (though ambitious) set of goals, it created the identity and momentum necessary for AI to emerge. It was the moment AI stepped out of theoretical papers and into the labs as a distinct, ambitious, and world-changing field of scientific inquiry. The ribbon was cut, the journey had begun, even if the destination proved much farther away than the optimistic pioneers initially thought [10]. $\epsilon$

References

1. AI History, Groove
2. Wiki, Oxford Academic, PhilPapers, SEP
3. ResearchGate, Open Encyclopedia, Minds and Machines
4. Aveni
5. Dartmouth, HoDS, Securing, AI Mag (PDF)
6. CHM, JM (CHM), Teneo
7. Datategy, Wiki
8. Wiki, CPW
9. Wiki, Quanta Mag, Seminal Paper (Wiki, PDF)
10. Stanford

Take a quick Quiz on things you've read: