My personal history as it relates to AI

Until the late ’90s I was a complete layperson. But at that point in my life after having worked outside (construction, landscaping, gardening, farming, etc…) following college I was looking to try to earn more income as our kids were getting older and we were looking towards the time when we’d have to support their college educations. I had worked in NYC for a few summers while in college and then for a few years afterwards but at the time (as a young active guy) I felt it would be soul-crushing to have to “permanently” work behind a desk. Hence the outdoor jobs (which I mostly quite enjoyed, but didn’t pay very well).

My mom gifted me and my family a PC as a Christmas present in 1995. I hadn’t had/used a computer for years prior, but persevered (having not “opened” the computer before X-mas morning and then getting frustrated for several hours (as the kids eagerly wanted to “play games” on it) before finally succeeding at installing Windows 95 and getting various video driver issues worked out on my own). I intentionally didn’t get a modem for quite a while (about 6 months) wanting to try to get familiar with the computer before connecting to the rest of the planet digitally. Eventually I did, and quite soon realized that GNU/linux (Debian in particular) would not put up the “barriers of understanding” which the opaqueness of the Microsoft Windows Registry (with its proprietary secrets) stuck in my way. But that’s not the story I want to tell now, so I’ll jump forward a few years.

As I said, college education for the kids was beginning to loom and I wanted to do more than the few side jobs I’d begun to pick up tutoring novices and fixing computers, connecting peripherals, etc… So I started to try to envision my “dream job” and (as I’m sure a lot of people have done and still do) I imagined working for a company helping to write, design and create video games. Little did I know that reality would far exceed those (seemingly quite lofty) hopes.

I was put in contact at that time(via fellow alumn Matt St. Rassler) with former classmate Ben Goertzel, who had just started Intelligenesis.net as a company trying to create artificial intelligence. After a few days of emailing back and forth, he offered me a job and I accepted. I commuted to the office below Wall Street in NYC 2–3 days a week and worked from home the other 4–5 days (it wasn’t like any other job I’d ever worked at and so the idea of taking “time off” just didn’t occur to me). I think I was employee number 17. While I’d had some experience with Pascal and Basic previously, I didn’t know Java, so getting up to speed on that language was one of my first challenges.

I started off coding (actually porting from _Recipes in C_) a fast Fourier transform (to be used for signal analysis as part of a larger toolkit in order to try to predict future (stock) market movements) and was then given the task of coding a wavelet transform. My French language skills weren’t up to the challenge of reading Ingrid Daubechies’s documentation of her (then fairly new) work, and the extra dimension (of wavelet series as compared to Fourier series (I very much had a solid understanding of the latter)) also really challenged me. So the code I attempted to write didn’t really work properly, but I was saved by the hiring of physicist Suzanne Yoakum-Stover, who whipped my code into shape (or maybe she rewrote it all from scratch? my memory is hazy at this point) and I enjoyed working closely with Suzi for the rest of my time there.

The endeavor blossomed (in the first dot-com era) and while swelling to over 200 employees (all over the world) and changing our name (easier for investors to spell (I always thought that was a stupid reason, but I was very fond of the original name)) to WebMind.com, we all set out to collectively change the world. It was a fantastic experience and I learned so much as we attempted to grow our codebase(s) into functionality which could accomplish many goals.

Later on, I learned python (which I like a lot more than java to this day) and worked on trying to get external sources of information, like CYC and WordNet into webmind’s architecture of nodes and links properly. I also spent a fair amount of time and effort on trying to help do the same with mathematical theorems.

But in the end, we became just another dot-bomb, evaporating in the spring of 2001. I’ve continued to remain in touch with Ben though (and he’s never stopped pursuing his goal of creating machines capable of true thought) while I worked outside the field (but still involved with technology).

I feel I have a view into how minds work due to having experienced my own breaking down completely.

I’ve had 2 full blown psychotic breaks in my life: the first in November of 2009 and the second in June of 2014. I use the qualifier “full-blown” because in the intervening years, I had some very “close calls” (none of which fortunately required hospitalization). While I don’t consider it the cause of my mental health issues, it’s obvious that the event which triggered my spiral down the rabbit hole of insanity was the death of my mother in a car crash in April of 2009. At that time I lacked the self-awareness that my subsequent struggles and treatment have made available to me today. All I knew was that I felt a terrible existential angst, and in the following 7 months, my thinking veered off track until it became completely delusional.

I remember dozens if not hundreds of specific delusions from the hours and days prior to my hospitalization in November of 2009. The tricky part about losing touch with reality has to do with the fact that it’s really hard to notice anything changing. The psychiatrist who was called to the ER on that early morning just prior to Thanksgiving in 2009, Dr. Schweitzer, who later treated me for a number of years, pointed out that I was confusing facts with beliefs at some point (probably early in 2010) and instructed me to attempt to do better at differentiating one from the other. I found that during that autumn of 2009 that my perception was that I was recognizing new insights that I had previously not been willing to consider or accept. At the time, I believed these fictions as strongly as (or perhaps even more than!) I had ever believed anything.

Looking back from the much more stable mental position that I have again attained, one of the things I’ve realized is that the “fresh look” at reality that was given to me at that time (in the late fall of 2009) did provide me with aspects of true insight. With regard to many historical figures (and probably many more less famous individuals) there has always been a fine line between genius and magic. To give just a single example from my own experience, I became obsessed with the idea of “true value,” and found our cultural reliance on “price” as a stand-in (at least in western industrialized societies) to be sorely lacking. The way I saw things (and to some extent still do), our attention to the economic cost of everything from labor to natural resources and beyond seen through the eyes of our human-centered market is inadequate and contributing to incredible destruction and general suboptimality in many dimensions, from preventing most of our citizenry from embracing a true sense of honorable purpose with respect to our work and lives, to the plundering of ecosystems which we won’t ever have the means to restore and causing the extinction of so many different non-human species from the face of the planet. These are all issues I’d considered prior to descending into this state of perceiving a “non-consensual realit”, but the emotions which surged out of control (especially fear for the future) brought a focus to my thinking which has never really subsided. While my vision at the time, of trying to transition away from a monetary economic system to a gift-based one may be more than I can accomplish as one single person among billions, this does not mean I’ve abandoned it entirely. I’m not completely certain what the lesson to be learned here with respect to creating artificial general intelligences is, but it has something to do with being able to walk that fine line betweeen fantasy and reality.

Perhaps one of my most important insights from what I experienced is just how necessary our beliefs are. What we believe provides a coherence that allows us to function. Even when I’m not suffering from delusions, I’m making a lot of assumptions and piecing together a worldview that allows me to try to “make sense” out of the world. I am convinced that if we’re ever to have any chance of creating true human-level artificial intelligence, that it will be necessary for such a mind to create its own internal model of reality.

As I lost touch with reality in 2009, I found myself focusing on symbols, seeing meaning everywhere and making mental leaps having to do with causal connections between events which were often unwarranted. In many specific instances I’m sure I saw connections which had no basis in reality as I latched onto what were mere coincidences and attributed deep meaning to them. But again, there is a fine line to be walked: I’m sure there were also things I happened to “notice” about the interrlated aspects of our world and societies which a “normal” person doesn’t ever even think about.

As I became overcome with emotions (again, especially fear — to the point it produced extreme paranoia inside me), one of the things I worried about was the approaching singularity, and the acceleration of thinking which would occur once machines become sentient and have the ability to create more (and better) machines. In my delusional state, I came to believe that this event was imminent. In what I have since come to understand is quite common with many who lose touch with reality, I also experienced delusions of grandiosity, thinking that I would be central to trying to prevent this dystopian outcome where machines would take over, leaving humans with no real options at all to affect and/or control our future destiny. There is an interesting anecdote about how my attempt to manipulate time (and eventual belief that I could cause it to flow in reverse) played a central role in disassociation with reality, but that will have to wait for another time for me to write about it.

In some sense, I think one way to interpret what I went through consists of the following: In my extreme distress over the sudden loss of my mother, I sought to create my own “vision quest”, which bore more fruit than I ever would have previously believed could ever be possible. The stability of normal life was removed out from under me and I mentally (and physically, though I haven’t concentrated on that aspect in my writing here) flailed around wildly trying to make sense out of what I suddenly perceived as chaos all around me. In “coming back to reality” in the months and years following this experience, I honestly believe that part of the “cure” has been to reassemble a “fiction” about my life and human society(ies) which discounts and outright ignores the true realities of our current unsustainable path towards a potentially horrific future with regard to environmental destruction, civil unrest and other looming catastrophes.

I am fairly sure (though one can never be absolutely certain) that the psychotic break I experienced in 2014 was for me the last I will ever experience. But I carry with me strong memories from these life-changing events and I’m hopeful that from a more stable and balanced mental state that I can now work towards helping to create a better future, both by contributing ideas and effort to various AGI endeavors and also devoting myself to other endeavors that I see as key to both humanity’s and the planet’s future.

While my day job has not involved AI for almost 2 decades at this point, I’ve nevertheless continued to ponder it. And I’ve done my best to keep up with Ben’s various endeavors, as he’s kindly looped me in to his thinking on all sorts of topics. And in February of 2018 I subscribed to the Common Model of Cognition email list and began paying even more attention to what a bunch of smart people are trying to foster with that effort.

I’ve been reading a lot about biology, complex systems, physics, language, consciousness and various other topics that I believe are relevant to the goal of creating AI. I’m very interested in how minds (especially my own!) work(s). And while I’m also trying to devote mental time and energy to other problems, I can’t help but be drawn back into this ever so interesting challenge that I first embarked on more than 2 decades ago.