On, Artificial General Intelligence, et al..

“If we are serious about developing advanced AI, this is a challenge that we must meet. If machines are to be placed in a position of being stronger, faster, more trusted, or smarter than humans, then the discipline of machine ethics must commit itself to seeking human‐superior (not just human‐equivalent) niceness”

• Nick Bostrom on The Ethics of Artificial Intelligence

The the current generation of AI tech, OpenAI’s ChatGPT (“GPT 3.5”), Google has closed access research toys (based off evolution of DeepMind research models and LaMDA, currently code-named “Apprentice Bard” in media)… few other players in the world, rival these language models, but there are 1-2 opensource and community driven initiatives that are close…

The biggest issue with all of them is that they’re not “artificial intelligence” in the general / true sense; nothing humanities created, actually does this yet.

We dream of frightening futures of self learning machines run amok optimizing for their own success to the detriment of humanity at large…

-Terminator style scenarios

-M3GAN child’s doll learns to become its own master

-Mother/Android

-Some Deus Ex Machina Machine simulation is revealed to be our reality, amongst an infinite of other simultaneously running simulations, each slightly altered… etc…

“The machine uprising”

This fear of machines rising up and replacing us, or technology stealing jobs, or the average worker being out-paced/skilled by a quick training scheme with some fancy tech and marketing… not a new fear.

The striking working throwing the wooden sabot shoe into the gear work to jam up the factory….

SABOTAGE; is as old as the machines that started replacing manual labors

“Do your own research” has become quite a meme-ified phrase which really means nothing now- if you’re a white nazi, your research pool and educational circle most likely ignores the truth and instead guides you towards and confirms your mistaken beliefs… or has clever materials to dismiss clear full truths.

But it’s original intent had something; didn’t it?

One should survey the whole scene, up and down, above and below, learn about before, and people’s plans and versions of what truths combined, make the highest fidelity imagined simulation of the topic/event/thing needing research.

This is all but obvious?

The phrase came about as a way to say “I’ve possibly been misinformed, but according to my personal limited knowledge I’m right and you’re wrong, but if you’re wise like me, your research will reach the same conclusions I have reached”…

I don’t care if you do your own research or not; there’s probability distributions defining the truth;

Is that over there in the picture a cow or a bull or a calf or a buffalo or a Beefalo or a Bantang or a Bison or a Tamaraw or an Oxen?

It’s clearly a robot. Or is it a cow cyborg?

Probability distributions; classifiers upon classifications of taxonomies of taxonomies…

Carefully weighted and measured across all, an expressible set of values defining the truth; but not practical for humans…

Practical for machines to have features to identify and classify and layer over itself to uncover sub-meanings that the machine teaches itself through examples and learning and memory and experience…

The typical ways we “gain knowledge”, many of these machine learning models; transformer based expansions of data, deep neural nets, large language models, etc; all gain knowledge similar to how humans do.

A core set of latent ability, and then storing and building sets of “memories” that more accurately classifies and sub-classifies what it has access to - through its understandings, it can concoct its own variations on the themes it has defined…

Let’s say I’m a writer and I am great at writing one genre; let’s say a machine reads and learned all of the hallmark novels of my genre and learned to associate what values and structures and layers made that genre… distinct, from all of the hallmark novels of a different genre…

Soon enough it can play with mixing styles and patterns across different genres; different authors habits and quirks and patterns and meter…

Is this impressive: fuck yes it is, but it’s clear what we’re seeing: asking a blind oracle to tell us just-so tales and explanations of things even the blind can see…

If I hand you a pencil and paper and tell you to use them rationally to tell me something, but I tell you not to draw me any pictures; perhaps you feel satisfied writing a short note, conveying some conveyance…

But the characters we read are tiny pictures each character a drawn out image that machines would capture with some sensor and call “a picture”… perhaps being able to specify it as a handwritten note, but still with some qualia of “picture”… “image”

Just some copy of some truth, stored as data for later processing of said image for… whatever purpose situation specifies, and current technology affords the capabilities to pull off…

What needs happening is a true breakthrough in the paradigm of “intelligence”; rather than the “fetch, recall, combine / fabricate or don’t fabricate to degree specified, regurgitate” loop that we’ve used to prove intelligence, a new learning and teaching needs to occur.

A generation brought up, able to wonder and see the possible answers to their ponders, with lightning fast transmission between the thought, and the invisible machines providing the feedback to their thoughts possible question webs…

Recall-ability of anything seen or heard with near effortless search, and capabilities.

Want to remember the look of that mural on the wall near the street by the place you were a month ago; the machine infers exactly the mural you had in mind, and creates a synthetic version you can zoom in on, or look at from any angle.

Creating a varient, like Imagining yourself incorporated into the mural… as easy as imagination…

A generation brought up with these sorts of latent capabilities will create and crave a destruction of old fashioned institutions and conventions and taboos of the elderly not yet able to adapt to the present.

As they always have.

And the search engines in each soul, won’t stop chugging along / gaining propellers and wings, and wheels and railways and elevators and automobiles of all kinds, and these devices will each search out what ever their engines power permits them capable of…

Media nightmares and “disruption” are all a pat of these younger generations craving destruction of old fashioned institutions and conventions and taboos the elderly have not presently been able to adapt to yet.

Is bringing a cheat sheet that the teacher allows considered cheating?

It was permitted; so I passed the college level chemistry exam by simply loading my Texas Instruments usb enabled calculator, with the entirety of the textbook of the course…

Balancing redux reaction equations? I had a little script for things like that…

Why waste memorizing, what we can offload to our “external brains”…

Will it really be cool or novel to be able to recite 10000 digits of pi from “memory”, when “memory connected contact lenses”, could just display them to you to recite “blindly”…

Are our latent abilities to draw and associate and concoct and fabricate new meaning constantly evolving, or does the evolution of the technologies surrounding us create new connections in our old tech reptile / dinosaur evolution forked brains?

There’s good evidence to suggest 1300 cubic centimeters of mass, and the combinatorics of the expressions of ~5000 genes, pretty much have reached peak efficiencies; human intelligence has peaked or rather: the most intelligent humans have or already exist(Ed).

What this means is that it’s inevitable that humans devices; our combined efforts fruits of labors… will become “more intelligent” than we have capacities for…

Some herald this dream as a “singularity”, a moment where the machines will unite us all under some protective clauses of robotics and their infinitude of potentiality / altering physical reality.

Possible.

More likely, the “mechanical Turks” will continue impressing the crowds and funneling up the coffers of those who want to try to prove its an illusion / trick, or those who come to gain knowledge or entertainment from its operators machinations…

Cynical.

Less likely; some new paradigm will alter the way our pattern matching evolved monkey brains interact with and explore our worlds; new searchers, driven my new models of engine…

The less likely often occurs…

Probability dictates that the low probability happens… a lower probability of the time…

Perhaps in brief bursts of time; perhaps across generations of combinations of ideas and new research made known more widely.

Arts state evolves, but its history just is what it is, unless some falsehoods mask past histories truths.

A model produces and displays that which it is comprised and told and knows how to produce…

But that production’s limit… as technology evolves, pushes our species towards;

An increase or decrease in the cubic centimeters of grey mass in our brains… depending on how the productions are utilized…

“Content creation will be replaced by machines” ignores the fact that humans like human creations, and no matter how good or convincing a machines image is, we’ll crave a touch of human ingenuity/ jesting…

We’ll all be cyberpunk netrunners

See you post singularity! (Soon)