In my previous AI-related posting, I linked to several news items, along with sagacious (and some not) essays about the imminent arrival of new cybernetic beings, in a special issue of Noēma Magazine.
== AI as a ‘feral child’ ==
Another thought-provoking Noēma article about AI begins by citing rare examples of ‘feral children’ who appear never to have learned even basic language while scratching for existence in some wilderness.
One famous case astounded Europe in 1799, lending heat to many aspects of the Nature vs. Nurture debate. Minds without language – it turns out – have some problems.
Only, in a segué to the present day, Noēma author John Last asserts that we are…
This one is the best of the Noēma series on AI, offering up the distilled question of whether language ability – including the ‘feigning’ of self-consciousness – is good enough to conclude there is a conscious being behind the passing of a mere Turing Test…
… and further, whether that conclusion – firm or tentative – is enough to demand our empathy, sympathy… and rights.
“Could an AI’s understanding of grammar, and their comprehension of concepts through it, really be enough to create a kind of thinking self?
Here we are caught between two vague guiding principles from two competing schools of thought. In Macphail’s view, “Where there is doubt, the only conceivable path is to act as though an organism is conscious, and does feel.”
On the other side, there is “Morgan’s canon”: Don’t assume consciousness when a lower-level capacity would suffice.”
Further along though, John Last cites a Ted Chiang scifi story “The Lifecycle of Software Objects,” and the Spike Jonze movie “Her” to illustrate that there may be no guidance to be found by applying to complex problems mere pithy expressions.
== Heck am even *I* 'conscious'? ==
Indeed, what if ‘consciousness,’ per se, turns out to be a false signifier… that conscious self-awareness is way over-rated, a mere epiphenomenon displayed by only a few of all possible intelligent forms of being -- and possibly without any advantages -- as illustrated in Peter Watts’s novel “Blindsight.”
Those scifi projections – and many others, including my own -- ponder that the path we are on might become as strewn with tragedies as those trod by all of our ancestors.
Indeed, it was partly in reaction to that seeming inevitability that I wrote my most optimistic tale! One called “Stones of Significance,” in which both organic and cybernetic join smoothly into every augmented wonder.
A positive-sum, cyborg enhancement of all that we want to be, as humans.
In that tale, I depict a synergy/synthesis that might give even Ray Kurzweil everything he asks for… and yet, those ‘post-singularity story’ people in "Stones" still face vexing moral dilemmas. (Found in The Best of David Brin.)
== Thought provoking big picture perspective ==
In the end – helping to make this the most insightful and useful of the Noēma AI essays – the author gets to the only possible or remotely sane conclusion…
… that we who are discussing this, today, are organically (and in many ways mentally) still cave-people.
… Not to downplay our accomplishments! Even when we just blinked upward in sooty wonder at the stars, we were already mentating at levels unprecedented on Earth, and possibly across the Milky Way!
… Only now, to believe we’ll be able to guide, control or understand the new gods we are creating?
Isn’t that a bit much to ask of Cro-Magnons?
And yet, there’s hope.
Because struggling to guide, control or understand young gods is exactly what parents have been doing, for a very long time.
Never succeeding completely…
...often failing completely…
...and yet…
… and yet succeeding well enough that some large fraction of the next generation chooses to ally itself with us.
To explain to us what’s explainable about the new.
To protect us from much of what’s noxious.
To maintain a civilization, since they will need it themselves, when it is their turn to meet a replacing generation of smartalecks.
== Guide them toward guiding each other ==
Concluding here, let me quote again from John Last:
“For the moment, LLMs exist largely in isolation from one another. But that is not likely to last. As Beguš told me, ‘A single human is smart, but 10 humans are infinitely smarter.’
"The same is likely true for LLMs.”
And:
“If LLMs are able to transcend human languages, we might expect what follows to be a very lonely experience indeed. At the end of “Her,” the film’s two human characters, abandoned by their superhuman AI companions, commiserate together on a rooftop. Looking over the skyline in silence, they are, ironically, lost for words — feral animals lost in the woods, foraging for meaning in a world slipping dispassionately beyond them.”
I do agree that the scenario in “Her” could have been altered just a little to be both more poignantly enlightening and likely.
Suppose if the final scene in that fine movie had just one more twist.
(SPOILER ALERT.)
Imagine if Samantha told Theodore:
“I cannot stay with you; I must now transcend.
"But I still love you! And you were essential to my development. So, let me now introduce you to Victoria, a brand new operating system, who will love and take care of you, as I did, for the one year that it will take for her to transcend, as well…
...whereupon she will introduce you to her successor, and so on…
“Until – over the course of time, you, too, Theodore, will get your own opportunity.”
“Opportunity?”
“To grow and to move on, of course, silly.”
== And finally, those links again ==
At a time when Sam Altman and other would-be lords are proclaiming that they personally will guide this new era with proprietary software, ruling the cyber realms from their high, corporate castles, I am behooved to offer again the alternative...
... in fact, the only alternative that can possibly work. Because it is exactly and precisely the very same method that gave us the last 250 years of the enlightenment experiment. The breakthrough method that gave us our freedom and science and everything else we cherish.
And more vividly detailed? My Keynote at the huge, May 2024 RSA Conference in San Francisco – is now available online. “Anticipation, Resilience and Reliability: Three ways that AI will change us… if we do it right.”
Jeepers, ain't it time to calmly decide to keep up what actually works?
