An Unscheduled Interrupt on AI
RAV’s Frank Gaffney discussing existential crisis of AI
with statecraft guru Sam Faddish. What could go wrong?
This, for one thing:
He has own gaggle of Chicken Little AI experts.
And, inevitably, this:
Me, when I’m not hiding my true self behind a photograph
They’re all worried about AI. Terrified in fact. For different reasons. Bannon has some apocalyptic seer who’s worried about meta-humans replacing Homo sapiens. Gaffney and Faddish are fixated on the latest AI hot topic, “The Singularity,” that magic moment when a supercomputer becomes self aware and decides to exterminate mankind.
How did we get here? The usual way. The movies. First, there was HAL:
Then, just a quarter century later, after a decades-long detour through a dead-end time paradox called The Terminator, there came the monstrous harpy HAL, a transgendered terminator in a box, like we’d been brought up to believe in, including Gaffney and Faddish:
Worried yet?
Here’s her resumé and mission statement.
Clearly, Ms. Hellz-A-Laughin’ here has had her moment of self awareness, that ‘Singularity’ which promises doom for all lesser intelligences.
The problem is, the Singularity is not a certainty. Indeed at the current level of our technology it is not even possible. This does not eliminate the real and increasing danger, which is the destruction precipitated by over-investment in a flawed technology and business application of the new software being developed.
Let me repeat: the Singularity described by (guess who?) Google’s AI information window is a fantasy, built upon another more established fantasy, that Black Holes Represent proof of concept for Singularities as a ‘strange attractor’ in physics. Also not true because not proven.
If black holes were true singularities, the universe would already be gone, consumed by the infinite density of some imploded star. If they exist (still unobserved btw), they are an extreme manifestation of gravity, but not an actual hole in the cosmos. The kind of Singularity affirmed by the AI true believers is in fact a resort to the oldest enemy of science, magic. It does not matter how many and how ambitious software programs are written, they are still written things, originated in some programmers’ minds. It doesn’t matter how big databases get, how many and how complex the computations they perform, they do not possess the necessary attribute to embody or even model infinity. Why? Because they are digital, not physical. They are an extremely limited subset of the capabilities residing within the human body and the universe itself.
The questions to be dealt with here (and they can be dealt with) are easily broken down by resort to a handful of enigmatic tropes.
1. The one about an infinite room full of monkeys typing on keyboards. Eventually, it is affirmed, one of them will, quite at random, type the complete works of Shakespeare. No, they won’t.
2. Opponents of it claim that Darwinian theory positing the creation of the vast array of living things, plant and animal, through a physics driven by random change and entropy, is akin to a hurricane blasting through an airplane junkyard and randomly assembling a completely operational 747. Won’t happen because it can’t.
3. The Turing Test, long touted as the litmus test for determining the achievement of self-awareness by computers, is like saying that a Las Vegas magician who convinces an audience he has made his assistant disappear from a locked box onstage has actually made her disappear. All it can ever be is a clever illusion.the Turing Test has never been anything more than a recipe for a hoax.
4. The arrival of huge new AI programming laid over top of the existing computer infrastructure should be invoking memories of the unforeseen consequences of incentivizing middle class homeowners to install solar panels on the roofs of their houses, what was supposed to eliminate energy bills forever very often resulted instead in the crushing of those roofs by the weight of solar panels they hadn’t been designed to support.
[There is no blank slate the AI boom can write on to remake the computer universe. I write my first computer program on the card punchers used to enter code into the mainframe computers that overwhelmingly dominated the computer environments of major corporations and universities. Think those antique Gibraltars of hardware and software are gone? Think again:
AI is already causing massive problems in the existing infrastructure at every level. The delusion that computer super intelligence is still worth the risk is delusional.]
Why am I so sure that AI self awareness is delusional? In the first place I’m sure because no computer program current or contemplated can do what I have already done here. AI cannot do this kind of writing. It can only recombine other pieces of what has already been written to imitate what a conscious writer does when he’s venturing into the unknown,
You can believe what I just wrote or not. I am, after all, the writer who has posed this is a first postulate of my profession.
Was I kidding? No.
Fiction is any recorded attempt by a component of the reality it is part of to represent its own version of that reality as the Truth. It is always a subset, always dependent on symbolic representations that leave what bleeds into infinity (i.e., everything) out of the record except by acknowledgment and reference. In this context, every kind of writing is fiction by definition, including poetry, plays, and novels, yes, but also legal briefs, textbooks, scholarly treatises on every subject there is, personal diaries, court transcripts, newspapers, shopping lists, posit notes, and love letters. And computer programs.
Getting closer to a possibly successful symbolic representation of reality is a function of how many tools of perception and communication the writer has. Regardless of its size and seeming sophistication, every computer program begins at the keyboard of the programmer. That he cannot think the program into exist but must type it leaves more than 99 percent of reality out of the creative process immediately. Keyboards generate digital information in the yes/no language of electricity, which has two states, on and off. Computers can model individual keyboard characters in code that assembles strings of distinct yes/no combinations symbolizing a, b, c, #, @, +, -, 1, 2, 3, etc. the code doesn’t know what characters are. It doesn’t know what numbers are, or words, or grammar, or paragraphs. Operations on characters and other coded data can be performed by hardware instructions called machine language, which also doesn’t ‘’know’ what it is doing any more than a junkyard car crusher knows what it is doing when it performs its prescribed physical function of eliminating Spence between one surface and another.
That’s a description of all computer programming in a nutshell. The car crusher analogy is as close as computers ever get to physical acts, physical experience. The computer chip that controls air injection into the cylinders of an internal combustion engine do not ‘’know’ they are feeding the explosion of gasoline that produces power for things called automobiles. Now think of what else digital data and hardware operating functions cannot “know. They have no sensory inputs at all. They do not see, hear, smell, touch, or feel emotions of any kind. Although they can be made to count it, they have no sense of time. The only ‘powers’ they possess are speed of operation and vast storage capacity for piles and piles of code in different categories capable of being used in various operations by machine logic.
Why AI is called Artificial Intelligence. Clever programmers can make programs simulate human-like activities, sometimes to an extraordinary degree. But the displayed output of any program is always a fiction because it is a written thing, even if there are programs which maintain other programs in specified ways and can generate new ones in accordance with the instructions of a programmer who has since died or moved to Barbados.
There is nothing in any of these transactions that can make the enormous leap to self awareness except in the mind of a writer like Isaac Asimov, Arthur Clarke, or Ray Bradbury. We call that dramatic license, a permission given to writers to use their imaginations for the purpose making readers suspend their disbelief long enough to be entertained.
Now I will take an unforgivable step that is nevertheless a fiction closer to the grand reality than the current state of willing disbelief in the populace at large has unconsciously accepted.
The only perspectives that matter with regard to the understanding and modeling of consciousness are self aware writers of fiction. The so-called profession of psychology dropped the ball completely on this variable in the last half of the 20th Century. That’s when B. F. Skinner invented behavioral psychology in order to sidestep the mysteries of human consciousness by replacing them with a computer-like model based on boxes and surveys and and self-fulfilling experiments in behavioral responses like the ones originated to study dogs and monkeys. Identify the behaviors, normal and deviant, then treat the deviant ones with drugs that minimize the unwelcome behaviors, very scientific approach that reduced the suspect vagueness of unprovable Freudian and Jungian theories about how human consciousness is organized and interacts with reality (i.e., the outside world).
Of course, psychologists weren’t the only victims of the extraordinary 20th Century ascendancy of science as an organized institution rather than a profession. The whole academic world also fell under the spell of the soft science determination to play at being hard science. Academic success in all manner of disciplines became possessed by the lure of surveys, calculations of correlation and cause/effect in large populations and subsets thereof.
By osmosis if nothing else, the humanities also devolved into science-like specialties. The past had somehow lost its stature as a repository of inherited wisdom, and those who had been the guardian of that tradition abandoned it for less challenging realms. The guardians of literature became specialists in championing the forgotten writers in the currently fashionable subsets of ethnicity, race, religion, and gender. Philosophers had already committed suicide decades earlier, as scholars like Immanuel Kant used deep dives into words and their latent meanings to destroy the meaning of all words and therefore meaning itself. Which paved the way for existentialists and the film careers of Ingmar Bergman and Ken Russell. What remained of philosophy an academic discipline dumped Plato in the trash and put up shingles in the practice of the Philosophy of History, the Philosophy of Political Science, and even the Philosophy of Science.
The subject of consciousness, what it is, and how it works was left to writers of fiction, who are inescapably hostages to their own experience of translating fragmentary memories of events, images, smells, sounds, and emotions into words capable of moving the heart and mind of a reader. No matter that they also tried to run away from the mystery of what consciousness is and how it relates to what is generally accepted as reality. They were, are, and will always be in the business of facing the slippery business of being conscious and discovering new ways of exploring it.
Sad that so many writers have taken themselves out of the fight with their Luddite response to the AI phenomenon. Unbelievably, writers of fiction still cling to their old typewriters or limit their use of computers as tools of a keyed entry tool that eliminates the need to retype subsequent drafts. Any writer who has grappled with the challenges of imagination in turning memories and curiosity into new worlds undreamed of by others not so engaged could attest to the sheer impossibility of transforming imitation intelligence into real understanding of the eternal quest for meaning. They’re not doing it because they’re otherwise engaged, these days with a passion for politics that is all they have left of their once noble search for meaning.
Why should you believe me? Because I have always been obsessed with the search for meaning. I have been very close attention to how the mind — i.e., consciousness — works for 52 years at least. I have been writing longer than that. I wrote commercially for many years in the high-tech world of American manufacturing corporations, and I made good money doing that. I walked away to write about the inner world of individual human experience, got myself banned from the American book publishing business by writing the truth as I saw it, and have continued writing for the last 30 years for free. More than 3 million++ words and many thousands of computer graphics for which I never made a dime of profit.
Today I am a one-off. I have paid my dues in the two worlds that matter the most to the approaching crises of Artificial Intelligence: fiction writing and computer technology. No one else has quite the depth and breadth of my experience in both these realms. My assessment of the real danger? That a load of big AI applications and data will make everything fall apart, slowly at first, and then when a tipping point is reached, catastrophically in terms of system failures and economic collapse. Not SkyNet but a serious hurt to every man, woman, and child in the nation.
You don’t have to believe me. Up to you. You don’t have to read what I’ve already written. It’s available right here, as it always has been, if you want to evaluate my thoughts in more detail;
The Artificial Intelligence Hoax
The Artificial Intelligence Hoax 2
The Artificial Intelligence Hoax 3
Why the Vaunted AI Boom is a Bubble
Year End Thoughts for 2025 — Personal Preface
Year End Thoughts for 2025 — The Big Stuff
A Shaft of Light in the Darkness — An EOY ‘25 Postscript
8086 (Blow thru the 2 warning screens; there’s no security risk, just a Wordpress glitch)
Yeah, I’ll put up another Kindle book one day. Too busy with other things for now.
Comments