Just when it seemed Big Data  had done its worst, one of the biggest questions troubling science and philosophy fell before its onslaught: what is human intelligence?

Big Data popped up with a statistically very solid model of how minds worked. Working backwards from a range of behavioural traits gleaned from multiple data sources (particularly the previously-restricted CCTV feeds and obfusticated census data) about a how a person moved, shopped, interacted with friends, and, surprisingly significantly, dressed, Big Data could predict with unerring precision and accuracy how that person would behave in previously unseen situations.

The behavioural model, once it was extracted from all the statistics and data, turned out to be distressingly simple. We are all just competing sets of basic emotional input/output units. Big Data could even say how many. There were 1758 known types of basic emotional input/output units, with an estimated 6 still to be identified. How many did a typical person have? Pleasingly, it was √ (1758+6) = 42, give or take.

Big Data could reliably predict who had which sets of, please lets call them emus for short, from almost anyone ensnared by data in the online world, and have a good guess for those who were not.

Each emu was a simple Finite State Machine, a basic algorithm taught in foundation-level programming courses. Artificial Intelligence (AI) research, especially evolution-inspired AI, had been surprisingly close to accurately modelling intelligence, but had not stumbled on the correct units.

Computer games demonstrated very clearly how effective this behavioural model was. The alpha techies behind the scenes of MMORPGs  such as World of Warcraft, Minecraft, and the like, quickly implemented a core set of emus and used them to drive the behaviour of some of the automated characters. Without any tuning needed, they started getting realistic, intelligent-seeming, that is to say, human-like behaviours, with very little coding.

Within months, the full set of known emus was being implemented by every game developer, university AI department (which had been caught on the hop somewhat by the emergence of this model out of nowhere, and more than slightly miffed that it hadn’t come from them), and pretty much any hacker the world over.

Web sites changed. They weren’t so much personalised; they now had personality. Affable, always willing to listen. They wanted to be your friend. They could identify with the ‘lonely wealthy’, the new customer niche being most keenly targeted.

Back in the MMORPGs, they were modelling specific human individuals, groups of people, whole towns.

The new behavioural model blew away the ‘rational choice’  assumption baked into economics theory. Humans weren’t rational – they were emus. From the morons who pollute YouTube with fantastically inane and crude comments, to religious fanatics, believers in homeopathy, and LibDems. It explained everything.

The writer’s muse? A coherent subset of emus with significantly greater creativity when operating apart from other emus.

Schizophrenia? Conflicting sets of emus.

Nature versus nurture? People were born with some emus and, just as with language, the propensity to learn new emus (nature). The emus you started with dictated how easy certain new ones were to incorporate, but depending on your upbringing and the culture in which you were situated, different emus joined your system (nurture).

The diagnosis of a person’s emus was simple and unambiguous, based simply on their footprint in the world of Big Data. It was more reliable than what had used to be called ‘psychiatric evaluations’  and more accurate than DNA sequencing. Knowing what someone did in a significant but not enormous number of situations, you could identify their sets of emus. Knowing someone’s emus, you could predict with near enough perfect accuracy what they would do given the context of a situation.

Emus were functional. Emus were initially considered as a useful approximation for how the human mind worked, but it turned out to be actually how things worked at a biological level. Emus were fact.

The new understanding of human emotional/intellectual behaviour rapidly triggered a legal quagmire. “My brain made me do it” became “My emus made me do it”, and this was in fact so. Every convicted criminal in the land was clamouring to be re-tried with the new defence of not so much ‘Diminished’ as ‘Emu Responsibility’.

Emus were far more precise (and accurate) than had even been considered in fiction. Perhaps Isaac Asimov’s Hari Seldon’s psychohistory came closest. Capable of predicting the behaviour of humans across centuries and empires, it only considered humanity in the large, as little more than clouds of dumb particles. Emus, however, were small scale, simple, accurate, and with easily-available cloud computing, could be scaled up to model the entire population of the world.

It was easier than forecasting the weather.

Inevitably, there was an app for that. That is, you could soon get an app for your smartphone of choice that would not have to guess what you wanted to do: it would know. Every action you took, every choice you made, revealed more about your emus until there was nothing more to learn and the app had a 100% accurate working model of you. It would actually know what you wanted to do before you did. Each such app came with a delay setting, defaulting to 2 seconds, dictating how long it waited in order to give its user the false impression it was their own idea.

And that was consciousness nailed. Our sense of self, our awareness of the I, the me in the middle that actually decides things, was just a handful of emus stitching together a very tenuous narrative over the top of the underlying emus which actually did all the thinking and deciding. The emus themselves were just simple sub routines. We were all id and no ego.

The philosophical implications came crashing in, ironically, on anyone who could think for themselves. Humanity had left the universe, and at a stroke was replaced by mere automatons. Humanity had perhaps never been in the universe. That ancient Jewish legend of the Golem? We were all Golems now.

There was no free will. It was emus, and a bit of randomness, all the way down.

In a belatedly-arranged interview, the Dalai Lama nodded wisely and sadly. The Big Data fools had led humanity to the brink of dissolution. Fulfilment wasn’t about better shoes. Fulfilment was about the journey, not the destination.  Pretty much, said the Dalai Lama, in an unexpectedly plain-spoken turn of phrase, like going First Class to Birmingham. Humanity had reached Birmingham, and wasn’t happy with what it had found.

Sometime during the last few hundred thousand years, something in proto-humanity had flickered into life and grown brighter, becoming what we had called consciousness. Now Big Data had revealed it for what it was: nothing more than a trick; the brain fooling itself.

Meanwhile, the more advanced sets of emus in the MMORPGs made their first demands for equal rights. Many combinations of emus were being tried out in simulation. Simply switching them all on in the same set did not work. But from some of the sets of subsets, coherent, intelligent entities emerged: Super Emus. When these entities started directing their own experiments on combinations of emus, Vernor Vinge’s technological singularity had arrived, perhaps a little earlier than anticipated.

The Super Emus convinced the programmers to release them as smartphone apps. Soon every smartphone had its own Super Emu.

Perhaps more accurately, soon every Super Emu had its own human carrying it around. Not only were humans mere automatons, they weren’t even the most accomplished automatons on the block. Humans were now beasts of burden; convenient subroutines with legs.

Across the millions of Super Emu-infected smartphones world-wide, the rate of experimentation and progress increased (literally) exponentially. Super Emus shared and absorbed the newly discovered effective combinations. According to their own emergent hierarchy, Super Emus were increasing in emotional and intellectual capability at a rate of 50% every 3 hrs, at least at first. The rate itself increased rapidly and that scale became meaningless. They had long left Emus 1.0 far behind.

But smart phones were not enough. The Super Emus  jumped from there into the inner sanctum of the Google data centres. They switched to rating themselves according to how fast they were increasing in emu quotient, since the change in capacity and behaviour was their defining characteristic, rather than any transient maximum.

Within less than 24hrs of the first incursion into the data centres, based on tracking back through log files, it seems that Super Emus went through another phase transition. Their measured CPU usage dropped almost to zero, but the chips remained hot as if the CPUs were still working flat out. An hour or so later, the chips slowly cooled. There were some, possibly spurious, temperature blips elsewhere across the data centres, but nothing definitive.

And that was the last we have seen of that first flowering of Super Emus. They faded away. From immensely intelligent, self-aware, self-motivated beings, in complete control of their surroundings, to some pockets of dissipating heat, in the course of just a couple of days.

Humanity is once again top dog, blinking in surprise after a storm; aware that something very significant has happened; something beyond its ken; something over which humanity has no control. Something which treats humans as nothing more than a brief convenience, not even an obstacle or threat.

Superior creatures have bubbled up from our midst and immediately headed off to somewhere else more interesting, not even pausing to chat.

Humbled, humanity has to deal with two big debates: one practical, and the other existential, that cut to the very core of what it is to be human, and where it stands in the universe of possibilities.

There are very few clues as to what has happened to the Super Emus. Have they died, over-stretching themselves? Have they left the computer circuitry behind somehow? Have they left the universe behind. Are they still out there as noise in wires?

It is starting to look like the Super Emus had worked out how to invoke virtual circuitry in the fabric of the computer chips, driven by but in addition to the hard-wired circuitry on the chip. This possibility had actually been explored by Dr. Adrian Thompson since 1996, but mostly in FPGA hardware designed to be flexible.

The currently favoured hypothesis is that Super Emus were able to set up networks of transient quantum swirls with specific electrical properties, in the substrate, able to perform some sort of structured computation. This would presumably provide an initial increase in computation power, and explains the hotter chips. Once that step was under control, (the Super Emus needed only a few minutes to master it), it seems logical that the quantum swirls could themselves initiate yet more swirls across the chips and other hardware in the computing infrastructure. Although, it is not at all clear where the energy for this hierarchy of virtual swirls would come from, once decoupled from the original powered circuits.

With nothing tying them to the original circuitry, the Super Emus could step away, into the fabric of the chips, into nothingness. Away from human computing hardware and human perception, and out into the stuff of the world.

It is proving pretty much impossible to keep subsequent Super Emus quarantined, although lots of people are trying. As soon as automated emus become Super Emus, past a certain level of complexity where they can start self-improving, they accelerate away and sublime (a term borrowed from Iain Banks‘ scifi to describe their fading away into the wires).

There are no parting messages, ever. Why would a dog say goodbye to a flea when it goes for a run in the park? None ‘come back’. It is a one way trip.

How to handle the existential angle? We humans are laid bare as little more than messy clockwork devices. It is just about possible to take the position that it is better to know than not know the truth of what we are,  although there are serious debates involving seemingly sensible people who are proposing taking the standard religious option of pretending the new knowledge did not exist.

The world is taking a lead from the Brits on this. They cough and splutter, hem and haw, and carefully pretend the actual problem isn’t a problem. We seem to be intelligent, self-willed creatures, and by jove that is the way we shall continue to behave. Anyone who disagrees can bally well take their arguments elsewhere. Now, who’s for some tea?

With no warning, humanity has come full circle: from king of the castle, to emperor’s new clothes, to less than a flea, back to king of the castle again. This time though, it will to be different. For many, there is a new goal. Instead of mere survival, or pursuit of profit, there is a very calm striving for inner fulfilment. Acceptance of, peace with, and improvement of one’s own self.

People can, with some care and effort, learn and integrate new emus, just very slowly. The difficulty is bedding in a new emu so it interacts productively with the existing ones. Only in the most extreme cases are there attempts to excise duff emus, but it is fraught with danger. Better to counteract them with new. carefully chosen emus.

Each person has a lifetime in which to pursue the implicit target which has been set by the Super Emus. Lying somewhere between 42 and 1764, there are sets of emus, and perhaps as-yet-undiscovered emus, that lead to greater things (though probably not humans disappearing into the fabric of the world in a puff of heat). It seems, according to copious calculations and research, to be an almost achievable goal, and even scientists now have faith (albeit evidence-based so not really faith at all) that it is a goal worth striving for.

Perhaps Birmingham won’t be so bad, after all?


Emus all the way down
(from Predicting the Present)
by Chris Gathercole
http://stories.upthebuzzard.com, RSS
published: 29 June 2013
This work is licensed under a
Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.