Dateline: 03/30/97
AS FAR as we know, human intelligence is the pinnacle of progress to date in the still-unfolding story of the cosmos. It is the most complex entity we know of, because it is the latest development in a process of complexification—of complexity heaped upon, or multiplied by, complexity.
In this article, I seek to set down my understanding of the key elements of that story as they pertain to AI, mainly to explain why this site at The Mining Company is organized and run the way it is. (As an unexpected spinoff, we also answer the most fundamental questions of all: Which is the world’s oldest profession, and Why are there so many lawyers?)
First, we look at the big picture of the fundamental scientific progress leading to intelligence, then we focus in on the specific modern disciplines and technologies that our intelligence is using to create artificial intelligence. Finally, we map the structure of the field of AI onto the structure of this site.
The Big Picture
In the beginning was the Word. And the Word was God, or Nothing, or Void, Tao, Singularity, or the laws of physics—your choice. Things were pretty simple, back then (about 15 billion years ago, if you subscribe to the Big Bang theory). Things don’t get much simpler than no things at all. Imagine a universe composed of just physics, math, and one very smart lawyer. And it’s timeless. Eek.
The lawyer figured the only way s/he was ever going to get paid was to introduce legislation mandating Time and Chaos, so s/he did. Given Time, Chaos could start to happen, and it did. With a Bang, part of whose energy was turned by the laws of quantum physics into a handful of quarks.
"Lo!" quoth the lawyer; "A brief!" Thus it came to pass that the laws of quantum physics further ordained that the quarks combine in various ways to form atoms of various sorts. The laws of chemistry ordered the atoms of various sorts to combine in various ways to form molecules of various sorts. Mechanical laws (later described by Newton) and space-time relativity laws (described by Einstein) ordered the molecules to shuffle around to form the stuff of space, of galactic swirls, stars, and planets of various sorts.
These laws themselves combined to order the formation of a pre-biotic "soup" on Earth. New and more complex laws were devised to govern the biological and evolutionary development of organic life of various sorts. Finally, laws we can barely discern, let alone fathom and describe, caused later evolutionary forms of life to assume culture (in which I include social, belief and value structures) and intelligence, of various sorts.
You, dear reader, are the current state of the art of all this progress and complexity.
Current Disciplines
Your biological intelligence is, in a sense, the product of the workings of the cosmological sciences, quantum and Newtonian mechanics, relativity, chemistry, biology, and culture, and hence of the myriad sub-disciplines that make up these huge fields of inquiry, multiplied by a chaos/complexity factor. But before we arrive at the creation or emergence of artificial intelligence, we have to add some newer sciences and technologies that could only exist because our bio-intelligence existed to create them.
Chief among these are computer science, cognitive science, and the technologies that both enable and support them. One of the most striking things you’ll notice is that they cross over into one another’s territory. For example, you’ll find linguistics under both computer science and cognitive science. So it's not just disciplines that are converging, but also—and more significantly—the theories and concepts they have studied and shared, sometimes under different names and usually from different perspectives.
For example, "computational linguistics" first arose as a concept within computer science/machine intelligence, but it is now considered (though not by everyone) to be a concept applicable to the way our own brains process language.
In short, the dividing line between biological systems and artificial systems is dissolving before our very eyes.
To provide a rigorous, or at least satisfying, account of the disciplines would take more space here than I suspect you are willing to imbibe in a single gulp. Therefore I will try to write about them individually in future articles. The following lists are just to give you an idea of the scope of the AI-related disciplines.
Computer Science
Computer Science is a catch-all term for such sub-disciplines as:
Algorithms, Architecture, Artificial Intelligence, Compression, Computer Engineering, Computational and Applied Mathematics, Computational Mechanics, Computational Learning Theory, Computer Vision, Databases, Distributed Computing, Computer Aided Design and Manufacturing (CAD/CAM), Formal Methods, Graphics, Handwriting Recognition, Human-Computer Interaction, Information Science, Knowledge Sciences, Linguistics, Logic Programming, Mobile Computing, Modeling, Networks, Neural Networks, Object-Oriented Programming, Operating Systems, Real-Time Computing, Robotics, Security and Encryption, Software Engineering, Supercomputing and Parallel Computing, Symbolic Computation, User Interface, and (phew!) Virtual Reality.
The above is not even an exhaustive list!
The key areas of study within computer science upon which I propose to focus most attention at this site are those I consider most proximate to machine intelligence:
The site's Resource List reflects this focus.
Cognitive Science
Like computer science, Cognitive Science defies precise constitutive definition. (It is often referred to in the plural, as "the cognitive sciences.") The Rutgers Center for Cognitive Science Web site provides one description of the field:
"Cognitive Science is one the few fields where modern developments in computer science and artificial intelligence promise to shed light on classical problems in psychology and the philosophy of the mind. Ancient questions of how we see the world, understand language, and reason, and questions such as 'how a material system can know about the outside world', are being explored with the powerful new conceptual prosthetics of computer modeling."
Essentially, the cognitive sciences are about the study of mind, brain, consciousness, and intelligence. Increasingly, computer science, neuroscience, and artificial intelligence studies are being subsumed under the term cognitive sciences.
The central disciplines are psychology, language acquisition and processing, linguistics, psycholinguistics, perception and attention, computation and linguistics, philosophy, visual sciences, and speech and hearing science.
The Center for Cognitive Science at SUNY (State University of New York) Buffalo, notes that
"… cognitive science has also been manifesting what has been a sea change in the direction of research in the social and behavioral sciences in this country: where previously the movement had been toward ever finer disciplinary distinctions, there is now the reverse dynamic toward an integration of the disciplines into a unified understanding." (Quote from SUNY’s Web site.)
This convergence and cross-disciplinarity is gratifying. I have been preaching it for years. But you know what’s amazing? You rarely see departments or colleges of sociology, communication, speech communication, or telecommunication included in the cognitive sciences. There is a smattering of anthropology here and there, but of communication—which you would (rightly) think of as a key discipline, there is little sign. So I’ll say it again: Communication scholars, get with the program! If you don’t, you risk becoming irrelevant.
Here is the logo for the Rhône-Alpes Federation for Cognitive Science. It illustrates the point that cognitive science is a multidisciplinary endeavor.
Besides the (absent) communication sciences, the key areas of cognitive science for AI from the perspective of this site are:
Again, this preference is reflected in the organization of the Resource List.
Technologies
The key and central technologies emerging from and contributing to the efforts of the computer and cognitive scientists are:
As I said earlier, it would consume more space than is tolerable (when reading from a computer screen) to discuss and describe these technologies within this article. We'll talk about ASR next week, and cover the other sciences and technologies in future issues.
AI at The Mining Company
This site is organized around these fields of inquiry. My goal is to bring you interesting and significant news of developments in any of them (including developments in our philosophical views about AI and its impacts) as such developments are reported, using the current weekly topic (This Week) as the vehicle. Over time, as more articles are written, the archive of topical features (Previous Issues) will itself constitute a reference resource for tracking developments in AI across the board of disciplines that go to make up the field.
A deeper and broader spectrum of the field of AI is to be found in the Resource List, where all of the more significant and/or substantial AI-related Web sites are organized by the elements (disciplines) discussed above, with the addition of specific areas for sites offering commercial AI products and services, and for human interest. The Best of the Net is an abbreviated resource list highlighting the very best of the sites listed in the major Resource List.
Finally, the Discuss area (at the top of the screen) takes you to a chat room and a bulletin board. In both, you get to have a pretty much unrestricted say, tempered only by the needs to stay on topic and to respect other participants and their views. I hope you will use these areas in part to help me develop and maintain the Resource List and Best of the Net.
There’s already a large number of AI related sites on the Web, and the number is growing. It will be difficult for myself alone to maintain these lists, and I’d appreciate your help in the form of email or bulletins to let me know of new sites, Web site address changes, and new content on existing sites. I would also be glad of feedback on the organization of the Resource List. If you think it should be structured differently, or would like to see categories added, removed, or modified, write to me.
Let this not be merely my soapbox. In the current and archived feature areas, I inevitably get to have the principle say in what gets published, but you can have a voice there, too. I’ll gladly share the soapbox with anyone who wants to contribute original thoughts or express well reasoned views contrary to mine. I must (if only out of my responsibility to The Mining Company) retain editorial control, which means I won’t give carte blanche to anyone, but if you have a valid and valuable contribution to make then you don’t need carte blanche.
Until next week,
NEXT WEEK: Automatic Speech Recognition. ASR is important because (1) it will provide the primary interface between homo and machina sapiens, (2) it is close to getting real, and (3) it will be the next major step in interface development, even without significant underlying intelligence. I'll tell you where the technology is at, and when to expect a competent ASR system on your PC.