'Accelerando' and the architecture of superintelligence

After four years I've just re-read Charles Stross's Accelerando and was again blown away. Here are my summer of 2013 thoughts. But in a nutshell, as the solar system transitions through the Singularity, posthumans of superintelligence transcend even augmented versions of ourselves.

Can we say more?

---

Intelligence is a reified concept. It presents as a trait of performance but is frequently conceived of as a mechanistic 'thing'. It's as if we took athletes such as Usain Bolt and Mo Farah and claimed they had a high 'quickness-quotient'. We would discuss the heritability of QQ and perhaps derive a scale. We would discuss biological correlates - is there a QQ module somewhere in the body?

IQ is clearly telling us something about performance, but to a greater extent than admitted it's collapsing several different things into one measure.

• Reaction time (where we're defeated by houseflies).
• Logical inference - where we're easily beaten by simple AI systems.
• Pattern induction (eg Raven's Progressive Matrices):  AI programs do well.

But perhaps the best definition of intelligence is the 'ability to learn or understand or to deal with new or trying situations'.

---

Back in March I wrote a piece about the architecture of intelligence, 'Roger Atkins: Mind Design notebook'.  I proposed that we should think of the intelligent mind as operating over a semantic network defined by nodes, links and overall processing speed.

-- Each node is a minitheory: some facts, rules and cached deductions + relevant inference rules. For example, you have a small minitheory about your pet and a much larger one about yourself.

-- Each link represents a kind of relation between theories (there are many). The classic 'ISA' relation familiar from object-oriented languages and ontologies would be an example. Also similarity relations for analogical reasoning.

-- Processing operations over a semantic net include:

• Take a node (minitheory) and deepen it with new facts, deductions or rules. Or create a new node on encountering or constructing some new entity.


• Create a new connection or relationship between nodes via an insight as to how they relate. A more richly-connected semantic net is potentially more powerful.


• Given a problem, navigate around the semantic net to form a solution (then add it).

---

With this architecture, a more intelligent entity has:

• a larger and more densely-connected link-set
• more and more-elaborated nodes
• faster link-traversal, new-link-creation and node-processing.

Links between different, and perhaps remote nodes will likely be rather abstract and removed from direct experience. For example, a notion of symmetry underlies both natural beauty and artificial design. A sophisticated semantic net requires the handling of complex abstraction.

---

This architectural model gives us a handle both on superior human intelligence and on posthuman superintelligence.

Firstly, why can't we just go and build an AGI today? Because the root set of competences in every human's semantic net are nodes and links which encode the experienced physical and social world, a net which requires a degree of innervated embodiment we have as yet no clue as to how we might build. Only when 'they walk amongst us' will designers be in with a chance.

Secondly, how would a superintelligence differ from today's humanity? A superintelligence would possess a semantic net with improved performance along all three dimensions. Observe however that no matter how complex a network of abstract nodes, at the base is the set of nodes which must connect to the complexity of the world. Even the brightest genius condemned to a sensory-deprivation cell wouldn't be that performative. Nothing there to work on.

I suspect that the sum total of new social experiences is parameterised by the possibilities of new physical environments, whether occasioned by exploration and/or technologies. And even these are ultimately bounded by the free energy available, as Accelerando reminded us with such gusto.

I think the take-home message is broadly as Stross imagined it. A superintelligence which walked amongst us (Hi Aineko!) would be bounded by the limitations of purely human technology + culture. But a society of superintelligences able to drive forward  their own physical, technological and cultural environment?

Well, let's just say there would be plenty of headroom.

---

Note: in this view, artificial neural nets are an engineering implementation of the semantic net architecture described above. We know from human neural nets (aka 'brains') that a 'compiled' semantic network runs real fast in the subconscious (maybe it is the subconscious) while trying to 'consciously' work on your own semantic network to address novel, complex problems is really hard work and a real test of IQ.

Smart pets? An *obvious* spinoff ...

From Steve Hsu today at Information Processing:

"Work on cognitive enhancement will probably be done first in monkeys, proving Planet of the Apes prophetic within the next decade or so :-)"

Many (most?) of the alleles implicated in superior cognitive functioning will doubtless have other physiological effects. This 'multiple and diverse impact' of genes is called Pleiotropy.

For this reason - when we have a list of the relevant IQ alleles - we would be ill-advised to shift into genome-editing mode and just flip all those genes to their 'enhancement' variants.

Naturally, there'll be loads of research first on mice, monkeys (expensive) and .. cats.

Link

Contra Wittgenstein, I may yet share my life with a truly biologica Aineko before I die.

---

(Nothing worse than being incessantly nagged by an entitled creature which believes itself master or mistress of the universe and which additionally possesses claws).

Accelerando; Stour Head

I have been a long-time Accelerando rejectionist. I guess I bought into all the reviews like this.

I describe a book that I admired immensely but didn't enjoy, a book that's impressive but not very entertaining. Accelerando is an exemplar. The imagination that concocted this future comes from the percolating skull of one of SF's most striking talents. But this is a classic case of idea fiction in which the ideas overwhelm the sense of wonder. While intellectually stimulating, there's no emotional connection. And the combination of reams of tech-talk and a dearth of relatable characters makes this a book whose appeal will be limited to a particular phylum of SF geek.

OK, so who wants to read very clever but soulless writing by a bright guy who's familiar with programming, AI, economics, quantum physics, cosmology, relativity, molecular biology, ...?

A lot of critics don't want to, that's clear. But on impulse I grabbed the novel from the library last week and eventually got round to it: and it's amazing! Here is the plot outline (from the same hostile review) - but it surely doesn't do the book justice.

The story presents us with three generations of the Macx family, throughout the full run of the 21st century, at the end of which nothing recognizably human exists any longer, and the solar system itself is home to thousands of space-borne processors in which billions of posthumans live in a uploaded state. Bodies are merely meat machines, obsolete at the best of times. In the first three chapters, Manfred Macx, working in the early part of this century as a "venture altruist," has come up with a way to liberate himself from what he terms scarcity-based economies. He comes up with one brilliant idea after another, and just gives them away to companies to do with what they will. This enables him to live anywhere free, off the goodwill of those whom he's enriched. While Stross makes a number of dire economic predictions that are entirely valid (even uncomfortably relevant in this day and age of two-dollar-plus gasoline), what I found interesting was an undercurrent of naivety in Manfred's ethos: to wit, that nothing he gets is actually free, though it's free to him. Somebody must still be paying for it.

Manfred's goal is to free the entire human race from outmoded economies and make everyone filthy rich so they can enjoy the Singularity when it hits. He also wants new laws in place to protect uploaded humans, which can only exist if old ideas about ownership, copyright, and intellectual property go the way of the dodo. When a group of lobsters (!) serving as test subjects for uploading somehow achieve sentience via the process, Manfred helps them to escape the Earth by beaming them in the direction of a signal emanating from deep space, thought to be of intelligent extraterrestrial origin.

This segues us into the middle third of the novel, in which Manfred's daughter Amber, fleeing from a nasty custody battle between Manfred and her neo-Luddite mother (via the clever use of new pre-Singularity corporate paradigms initiated by Manfred) ends up in Jovian orbit amongst a group of orphans working to decode and respond to the alien signal. This section of the book is where some structural problems blemish the plot, possibly related to its compiled-novellas nature. When Amber's mother tries a tricky legal maneuver to get her back, Amber — within about the space of a page — sets herself up as empress of something called the Ring Imperium, and promptly sends uploads of herself and the rest of her crew in the direction of the alien signal.

Amber should have been a fantastic character, but I could never warm up to her. And her section of the book, despite being no less suffused with spectacular ideas than anything else from Stross's pen, was one I simply couldn't get into — suprising, as it contains an amusing riff on first contact tropes and one or two pointed spoofs of 2001. (Not to mention its last chapter, "Nightfall," was a Hugo nominee.) The final section, involving Amber's son Sirhan, is interestingly the most engaging of the book, considering that it takes place in the twilight years of the century, when only a few "meatbody" humans remain in the solar system. In some of the story's most trenchant observations, we see that posthumans have much the same petty problems as their flesh-and-blood precursors. A dysfunctional family is pretty much the same, whether living today or in the accelerated future.

Actually the characters are interesting and their plot-relationships draw you in: it is a page turner. What I was truly impressed by though was the care with which Stross has really thought about the Singularity. Suppose the process of runaway intelligence was via the conversion of more and more of the 'dumb matter' of the solar system into "computronium" (i.e. dust-sized distributed processors). Why wouldn't a new Darwinian paradigm emerge? Why wouldn't the post-human intelligences cluster near the sun for power and to minimise speed of light delay. Why wouldn't it all go terribly wrong? So like everyone else, I'm blown away by the torrent of worked-out ideas, and swept along by the story. It's time to read some more stuff by Charlie Stross.

---

We spent the afternoon today at Stourhead. Clare was determined to give her mitochondria a thorough airing and set off relentlessly from the "Spread Eagle" complex of shops and cafes, through the gardens and woods to the far end of the estate near King Alfred's Tower (four and a half miles uphill and back).

I marvelled.

Here are some pix of the lake and gardens.

Clare appreciating the honeysuckle

The ducks - that's grass they're eating?

The author occluding the Rhododendrons