What even is… AI?: A retrospect 2: The Electric Boogaloo

Posted by:

|

On:

|

I did it. I made the joke. I said I might do it.. and then I did it.

I’m proud of myself.

Anyways, where did we leave off last?

Oh yea.. this guy:

This brilliant, quirky, fantastic bearded man, the great John McCarthy. If you check the last post, I provided a little introduction and link to some details about em’, but let’s dig into it.

Remember when I said at the time, it was INCREDIBLY expensive to think about computers, being used in mass, by normal people? It still kinda was when John came around, but there at been some proof of concept work that was initiated by three men and a baby.

I’m kidding.

Cliff Shaw, Herbert Simon, and Allen Newell (random order of names, not an order of contributions) generated a more in-depth proof of concept via the “Logic Theorist.” This was a software program that was designed to copy the skills human beings have when solving problems. Like… in 1956. Imagine that? At a time when politics weren’t greater, a huge number of people in America were being marginalized and oppressed, three people took a computer and used it to prove theorems in symbolic logic from Whitehead and Russell’s Principia Mathematica. What’s that you might ask?

“”Principia Mathematica” is a landmark work in the field of logic and mathematics written by Alfred North Whitehead and Bertrand Russell. Published in three volumes between 1910 and 1913, the book aims to provide a solid logical foundation for all of mathematics.

The work is notable for its attempt to derive all mathematical truths from a set of axioms using formal logic. Whitehead and Russell sought to show that mathematics could be reduced to logic—a project known as logicism. Their approach involves detailed formalization and rigorous proofs to ensure consistency and clarity.

“Principia Mathematica” is a landmark work in the field of logic and mathematics written by Alfred North Whitehead and Bertrand Russell. Published in three volumes between 1910 and 1913, the book aims to provide a solid logical foundation for all of mathematics.

The work is notable for its attempt to derive all mathematical truths from a set of axioms using formal logic. Whitehead and Russell sought to show that mathematics could be reduced to logic—a project known as logicism. Their approach involves detailed formalization and rigorous proofs to ensure consistency and clarity.

“Principia Mathematica” is famous for its complexity and depth. It introduces the theory of types to avoid paradoxes in set theory and explores various foundational issues. The notation and methods used in the book have had a profound influence on the development of logic and the philosophy of mathematics. Despite its ambitious scope, the project had limitations and faced challenges, such as the discovery of Gödel’s incompleteness theorems, which showed that any sufficiently complex formal system cannot prove all truths about arithmetic. Nonetheless, “Principia Mathematica” remains a crucial work in understanding the foundations of mathematics.

See? Fancy math and stuff. REAL HARD LIKE. Wanna know what’s cooler? The quote you see above was generated when I asked a Large Language Model to describe what “Principia Mathematica” was. This is what it gave me. Can I trust it fully? Probably not, but it gave you the gist of why it was so monumental when The Logic Theorist came out.

The question I always had after reviewing and reading through this program was… how come now one freaked out? Why wasn’t there world-wide panic? It’s 1956. A computer program was answering complex symbolic math. Not a single peep. Seems strange to think about, but I digress.

Back to the stuff.

The Logic Theorist was funded by a fancy place called the RAND corp, yes THE RAND corporation, and was presented at this BIGGGG conference (and by big I mean small, and by small I mean “It was ahead of its time” and what not, called the DSRPAI, or The Dartmouth Summer Research Project on Artificial Intelligence.

Thats a lotta words. Also, what’s with all the acronym in the 50’s and 60’s? Ain’t nobody got style back then?

Anyways. This Dartmouth conference was hosted by our bearded wonder John McCarthy, and another fellow Marvin Minsky. Minsky co-founded the Massachusetts Institute of Technology‘s AI laboratory, which is a very important aspect of Artificial Intelligence and its history. See McCarthy wanted to bring together a huge amount of people, and collaborate and discuss artificial intelligence as a new field of study. Heck, this was the conference where he coined the term “Artificial Intelligence”.

which by the way, I find so wild. HE INVENTED A STRING OF WORDS. NOW EVERYONE SAYS IT. They use it when they write about it, they use it when it’s studied, all that. I’ve used it 9,999 times within this post! (likely incorrectly) That’s pretty cool if you ask me.

Ole John was hoping that if he brought together the right researchers, then he could really flesh out the field, and focus effort on moving it forward in strategic ways. However, it didn’t go so great. People milled about, they discussed in groups, but they didn’t outline pathways for research, standard methods for the fields, or even what to do as a follow-up conference. While it didn’t hit like John wanted, they did all come to the same conclusion:

Artificial Intelligence is possible, it will happen, and one day we will have a sentient being that mimics humanity, but is inorganic in nature (or maybe a combination of both? I dunno, I’m not a futurist).

At the time, this conference didn’t seem like the second coming. Instead, it felt like a normal, valid, science-based conference based on research ideas. I don’t think anyone at the conference would realize that the talks they had, the conversations they created, and the terms they used would have a meaningful impact all over the world.. forever.

So what happened next?

Well… lots of things. From 1956 to the mid 70’s computers exponentially grew. They became faster, they became smarter (in the sense of processing power, and memory) but most of all they got cheaper.

Remember, before the Dartmouth conference, most studies or research fields in computers focused on computational physics. Tools for building environments to test or measure theoretical physics ideas. The term “Computer Science” had been around for a while, but it fully became a real field of study when the Dartmouth conference also coined AI. They were intertwined. AI was a subset of Computer Science (one of many subsets) and it took multiple innovations in many different subsets to push AI forward. By the late 1960s and early 1970s, computer science had become an established and distinct academic discipline, with its own set of theories, principles, and methodologies. The field has since expanded rapidly, encompassing various sub-disciplines such as algorithms, data structures, artificial intelligence, and software engineering.

side note: In relation to computer science, about a decade after the Dartmouth conference was one of the most interesting things I’ve read about in my time working through Computer Science Academia, called “The mother of all demos”. The “Mother of All Demos” refers to a groundbreaking demonstration of computing technology given by Douglas Engelbart on December 9, 1968. Engelbart, a pioneering computer scientist, showcased his vision for the future of interactive computing in a presentation at the Fall Joint Computer Conference in San Francisco.

Key highlights of the demo include:

Networking: The demo used ARPANET, an early packet-switching network, to connect multiple computers, demonstrating the potential of networked computing.Key highlights of the demo include:

Networking: The demo used ARPANET, an early packet-switching network, to connect multiple computers, demonstrating the potential of networked computing.

Graphical User Interface (GUI): Engelbart introduced the concept of the graphical user interface, including windows, icons, and menus, which would later become fundamental to modern computing.

Mouse: The demonstration featured the first public showing of the computer mouse, a device Engelbart and his team had developed, which allowed users to interact with a computer more intuitively.

Hypertext: Engelbart demonstrated hypertext, a method of linking information that allows users to navigate between different documents. This concept laid the groundwork for the development of the World Wide Web.

Collaborative Tools: The demo also showcased real-time collaborative work using networked computers, including features for shared workspaces and online discussions, which foreshadowed modern collaboration tools.

Word Processing: Engelbart presented an early form of word processing, which included features such as text editing and formatting.

Graphical User Interface (GUI): Engelbart introduced the concept of the graphical user interface, including windows, icons, and menus, which would later become fundamental to modern computing.

Mouse: The demonstration featured the first public showing of the computer mouse, a device Engelbart and his team had developed, which allowed users to interact with a computer more intuitively.

Hypertext: Engelbart demonstrated hypertext, a method of linking information that allows users to navigate between different documents. This concept laid the groundwork for the development of the World Wide Web.

Collaborative Tools: The demo also showcased real-time collaborative work using networked computers, including features for shared workspaces and online discussions, which foreshadowed modern collaboration tools.

Word Processing: Engelbart presented an early form of word processing, which included features such as text editing and formatting.

I’m cheating again. This top quote I pulled from chatGPT, because I wanted to make sure I didn’t missremember some of the things that were shown at the demo.

But isn’t that nuts?! It’s 1968, and they are talking about Hypertext, word processing, FREAKING GUI!? Graphical user interfaces.. in 1968. If you’ve never heard of him, I implore you to read up on Douglas Engelbart, what an absolute mad lad.

SORRY FOR THE DETRACTION…. BACK TO AI.

After the Dartmouth conference, and the work provided in “The Logic Theorist”, Newell and the boys (Newell and the funky bunch? Newell and his stable of bros? idk) worked on an additional extension of the original program called “the General Problem Solver” (sidenote: Newell clearly wasn’t focused on marketing cuz SHEEEESH). The GPS (more acronyms) aimed to solve a wide range of problems using a more general approach to problem-solving and heuristics. It was a significant step toward creating more versatile AI systems. It was less about following specific theorum proofs, or individual symbolic logic structures, and more closely aligned with how normal humans attempt to solve problems. While not a GIANT leap, it was a valid incremental step, and showed that AI could continuously get better and smarter, the more effort and research was placed on it.

You can read more about Cliff Shaw, Herbert Simon, and Allen Newell here.

So they’ve got a field of study. There had been incremental additions to the literature (I always laugh when I say that, but as a DCS student, “adding to the literature” is incredibly important when generating a dissertation), programs had been created to mimic human problem solving, and governments, organizations, and scientists were taking Computer Science (and by relation) AI incredibly seriously. So what happens next?

I’m sure there were a ton of amazing things between GPS and the next thing I want to talk about, but I think for historical BIG jumps in AI history, focusing on BIG additions is important. If you feel like something fits in between GPS and this, let me know. Maybe I’ll add it. Maybe I’ll act like it was my idea in the first place. Maybe I won’t even read your comment.

Who knows.

Anyways… I give you… by the sweet sweet grace of “super fun names”….drumroll please

THE PERCEPTRON

The perceptron has to be the greatest name for a Computer Science concept I have ever heard. It’s like when Steve Jobs and Steve Wozniak decided to use “Apple” as the name of their computer company. I just can’t get over how cool this sounds.

The perceptron is an early neural network model designed for pattern recognition tasks. The perceptron was an attempt to simulate the way the human brain processes information. It was introduced as a viable concept by Frank Rosenblatt. I first read about Frank in one of my favorite books by Steven Levy, called “Hackers”. It includes a bunch of great stories around the history of Computer Science, computers, processors, microprocessors, all that jazz. I probably read it once a year, it’s just so great. You should check it out.

So whos Frank Rosenblatt? Since we’ve been focusing on AI and Computer Science, you’d assume he’s a software engineer, or a MIT fancy computer priest.

You’d be wrong.

Technically, Frank was a psychologist. And I find that so amazing, when you think about his contributions. While John McCarthy is the godfather of AI, Frank is the big daddy of Deep Learning. Here he is, in all his nerdy perfect glory:

Frank Rosenblatt

So now you’re gonna say “but haxk, we’re talking about AI, not Deep Learning.. whats that? it has nothing to do with this story”

yea yea yea, I get it. AI is a lot of things. It’s a field. Fields have sub-fields. There are sub-fields of sub-fields.

ITS TURTLES ALL THE WAY DOWN.

But it’s an important distinction. Most of what we call “AI” today, is not AI. It’s not close. It’s generative, it has bits of reasoning, but it is not a direct copy or mimic of human thinking and decision-making. It’s a piece of AI, a particle of it. It will take a culmination of all of the subfields of AI to make AI.

But you want to hear more about Frank? He did some cool shizz.

FIND OUT NEXT TIME ON COLOR ME GEEK

sorry. I had to do that. I’ve heard that phrase “NEXT TIME, ON DRAGONBALL Z” so many times in my life, I had to do it once.

Or twice.. more maybe.. who knows.

But will continue in the next pose, talk more about Frank, and continue our walk through the history of AI, up until today.

All the love,

Haxk

Posted by

in