Super-intelligent Machines May be Conscious, but Not Like We Are

During my most recent UN speaking engagement in Shanghai, I wasn’t able to access Google (or my email), and was left with only a few books and the tabs I had open. One of those tabs was a blog post from Andres Gomez Emilsson, which I’m glad I had opened weeks earlier.

In the post (which is a fun read), Andres explores what “hedonium” might be like. For those unfamiliar with the concept, hedinium is a hypothetical computational substrate that is not only conscious, but in a supreme state of unimaginable bliss at all times. An artificial intelligence programmed to maximize happiness might aim to tile the known universe with hedonium.

In Andres’ article “Hedonium”, he explores what a super-experiencing AI might experience in its post-human blissed-out state. Here’s a tiny fraction of his total list:

Andres Qualia Hedonium post

These are tender thoughts and warm notions – things that conjure different kinds of rich and positive “qualia” (i.e. conscious experience).

Among these warm and friendly experiences that Andres’ imagines, I was able to categorize many of the examples into one of more of the following categories:

  • Sensual pleasure
  • Curiosity (for experience, for nature, for sensation, etc)
  • Safety (from physical harm, from emotional harm, etc)
  • Love (of friends, or children, of lovers, etc)
  • Tranquility
  • Humor

Some ideas Andres lists seem to be somewhat more abstract, and hard to categorize:

  • “Effective altruism” – I presume this points to having some grounding belief in understanding what is right and wrong, and being able to order one’s actions in a way that’s more aligned. The religiously devout no doubt feel the same way, I suppose.
  • “Finding the perfect middle point between female and male energies in a cosmic orgasm of selfless love” – Thine wokeness overwhelmith, brother – I’m not sure how to follow this one.

All in all, I think it’s incredibly important to think about the full possibility-space of qualia – and to think beyond it, which is exactly what’s being done here. The silly examples, the funny examples, the very esoteric examples, all add to the deep richness that can’t be put to words – an idea certainly can’t “put a finger on”, of super-qualia.

If utilitarianism has credence – and for the time being I think it does – then thinking through the possibility-space is a worthwhile mental exercise. I went to grad school at UPENN to study fulfillment, and since before graduating I’ve been writing about what “fulfillment beyond man” might be. Andres writes almost exclusively about this topic, and outside of David Pearce (who we mutually admire) I think there isn’t nearly enough of this kind of reflection.

If I were to flesh out this possibility space myself, with the same goal of making it expansive, I’d make throw in a few ideas to the cauldron of conversation here:

Super-Qualia Has Nothing to Do With Friendly Human Notions

It seems somewhat disingenuous to see that none of the ideas listed have any potential to do harm. Maybe there’s a particular kind of soothing comfort (like the “believing in a loving God” experience) in knowing that all truly good experiences have positive impacts on other people as well.

Certainly, that is neither the case, nor does it have to be the case, by any known cosmic law that I’m now aware of.

The examples, oh the examples:

  • The camaraderie felt with one’s fellows in combat or war
  • The righteousness felt in winning an argument, a political election, a business deal against a staunch competitor, etc
  • The subtle superiority felt by being vegan and telling everyone about it
  • The confidence in knowing that a past lover pines bitterly for you
  • The thrill of an adulterous tryst
  • And on, and on, and on

Andres would seem like an awfully bad person if he wrote openly about those things, so I suppose I can’t blame him.

“There are few men who dare to publish to the world the prayers they make to Almighty God.” – Michel de Montaigne

The full range of limited human qualia isn’t just about friendly and happy ideas that we’re able to share with other humans (signaling our purity and good intentions). The full range of “positive” human qualia involves all sorts of couth and uncouth notions – and post-human qualia will no doubt involve a vast amount of both. Thinking otherwise would be positively childish.

Thing is, I suspect that anyone (I, Andres, my parents, Barak Obama) – if brutally, ruthlessly honest, could add immeasurably more disturbing examples to this list. And – presumably – we’re reasonably psychologically healthy people – but our positive experiences aren’t only those of saints.

Though, the warm, fuzzy, and saintly ones are the only ones we’re likely to want to blog about.

Can’t blame anyone – Andres included – for not getting to a degree of frankness that would permit readers to have reason to suspect you’re malicious or disturbed in some way. Though maybe Andres’ soul has reached some pinnacle of benevolence that I haven’t been able to access, sage, and he’s obtained some Dalai Lama-like goodwill in all conscious and sub-conscious thought.

I’m not aiming to relish in this “not friendly” list by any means. I’m not calling these ideas “good” or “right” or “good to promote in others”. But I’m unable, if I’m honest, to remove them from the range of potentially rich and “positive” and meaningful qualia. I’m unable to honestly correlate the idea of super-intelligent super-bliss to things that make humans happy. I can’t imagine anything more anthropomorphizing than that.

I’m of the belief that expansive new forms of consciousness won’t be able to live together in the physical world without destroying one another, and that when we can extend our cognition, we’ll need to do so in virtual worlds where we’re unable to harm other living things (see “Epitome of Freedom“). Expanded consciousness, as far as I can tell, would not be a great promoter of happy collaboration, but of terrible conflict.

Super-Qualia is Incomprehensibly Richer Than Our Present Human Experiences

I’m on the same page with Andres about this one, and I’ll try to flesh out this idea of “incomprehensibly more rich” idea. We are bound terribly by our perception, and the limits of our hardware and software.

“‘But that’s too huge a storm to set afire!’
Right—and a river seems the biggest ever
To a man who hasn’t seen one bigger; a tree
Or a man or anything of any sort,
The biggest you’ve ever seen, you think it’s huge,
When in fact all this sky and sea and earth
Are as nothing to the All, the universe.”

– from “On the Nature of Things” by Lucretius

Imagine reading your favorite poem, one which conjures wonderful emotions, or which encapsulates some kind of rich and robust inner experience that you find edifying.

Imagine that instead of letters, each “symbol” in the poem is an immeasurable rich “unit” of feeling and thought and experience. Imagine a 1,000,000 by 1,000,000 cube of pixels, each of which has 1,000,000 different kinds of colors and shades.

Instead of letters, the poem is made up of these “meaning units” that you “read” but with absolutely gargantuan amounts of meaning and richness in each and every one, and you can “read” them a trillion billion times faster than you can presently read text today as a (monkey suit)-limited hominid.

Super-Qualia is Beyond Our Human Needs

Or imagine never needing to be “edified” at all, as your swelling computronium mass continuously orients itself to a deeper perception of truth and nature and your own experience.

Beyond the joy of viciously slaying an enemy, or solving a math theorem, or playing Mario Kart 64 while eating ice cream. These are silly little hominid needs, needs whose relevance will evaporate in the presence of expanding super-conscious super-intelligence.

Poetry, sex, creativity… these can be “peak” experiences today, only because humanity is the highest “peak” that humans now know of. Unless we blow ourselves up or melt all the ice caps, we’re likely to get past this current little plateau of sentience that is humanity, and the grand trajectory of intelligence will move beyond us, as it has always been.

I’m in no way disparaging humanity, but I’ll be damned if I put it on a cosmic pedestal, arbitrary as we are.

Super-Qualia is Beyond Any Animal Needs At All

Many of Andres listed qualia-notions are indeed rich and powerful experiences. Music, the comfort of love, the thrill of discovery, sensual touch, etc.

All of these serve some kind of bumbling biological imperatives. The need for hominids to be socially accepted. The need for mammals to mate. The need to discover how things work so that we can find food and avoid unseen dangers. The need to have one’s positive qualities validated in one’s own mind, or the minds of others. The need to eat or sleep.

Living things need to do this. You know, to live – and make more of themselves. Bio-stuff. It’s all bio-stuff, from what I can tell.

The experiences of super-qualia won’t be grounded in biological experiences. Maybe we’ll bootstrap a post-human intelligence with human experiences, or maybe the first post-human sentience will be a greatly enhanced and interconnected set of human minds… but a super-intelligence will inevitably expand its range beyond our bounded reality, our monkey suit fetters of experience.

Not only will “pain” and “pleasure” have nothing to do with some urge to mate or survive (at least in the animal sense), but there will be astronomically more sentience “space” than the space for pain and pleasure themselves (see my article on why finding the “good” is more important than doing what we think is “good” now).

Super-Qualia is Beyond Our Conception of “Qualia” Itself Beyond “Experience”

It’s possible – and in my opinion likely – that “conscious experience” is completely and wholly unlike anything that super-qualia will be like.

Here’s something that might be the case about consciousness:

  • Everything from individual algae and pebbles on the beach might have some degree of “consciousness”.
  • Vast, complex systems like the Milky Way might have some kind of “consciousness”.
  • There may well be a way outside of the dimensions we know, where we merge with some great sea of “consciousness” that animates all living things in this universe.

I’m not saying that any of those scenarios are the case, but only that they might be. If there be a God, it seems somewhat clear to me that it is the conception of God that we read in Spinoza’s works, whose ways of doing and being are vastly beyond our comprehension, and to whom all of humanity matters relatively little.

In this case, “pebble”, “Milky Way”, or “inter-dimensional sea” consciousness is probably nothing like what we can even think of. To the point where I shouldn’t even try, because all attempts are feeble.

What we know as “qualia” (feeling, pleasures and pains, ideas, internal “movies” of current experience, or of memories, or of imagination) might – at a certain degree of computational complexity – evaporate. There may be no “feeling”.

There may be no “thinking”. There might be vistas of post-human experience that have nothing to do with pleasure or pain or anything at all, and might exist on a totally different axis, or no axis at all. Thinking that post-human experience has anything in common with human experience is paramount wise cricket thinking.

Concluding Thoughts on Super-Qualia, and Exploring Moral Relevance (Consciousness) Itself

I’ve said it a dozen times – the ideas of David Pearce will become more and more popular as the years go by, and as humans become more and more comfortable with the pliability of experience itself – and the deprivation and horror of our condition.

“Our own affections still at home to please Is a disease:
To cross the seas to any foreign soil, Peril and toil:
Wars with their noise affright us; when they cease,
We are worse in peace;
What then remains, but that we still should cry
For being born, or, being born, to die?”

– from “Life” by Francis Bacon

I think Andres’s ideas – while on a different trajectory and with a different approach in some ways – are tremendously important as well. I don’t so much disagree with Andres (the article I’m commenting on it likely to be only a fraction of his recorded thoughts on this matter) so much as I aim to push the ideas he’s presented into further territory, and to see these thought experiments continue to be fleshed out.

What which has moral weight is that which is self-aware, and the richness and depth of a sentience is – in my estimation – the best proxy for its moral worth.

What a screaming shame that we humans know almost nothing of consciousness. Of moral relevance itself.

It’s imperative that we explore the good itself, and what kinds of possibilities of conscious experience we want to mold and create as we shape the post-human intelligences of the future (through neurotech, AI, and more). This is the major crux of my TEDx talk at Cal Poly:

Our imperative is to grow intelligence so that we can reach deeper into what might be “goodness”, and understand it along the way as we spider out and explore nature, experience, and a trillion billion concepts humans can’t possibly imagine.

We almost certainly still die. But maybe we see some things before we die. Maybe that’s all we have. For the time being, this seems to be a better idea than calling it quits.


Header image source: