Monday, 13 June 2016

The AI-Singularity

The following is the unedited version of an exposé currently featured in the 18th revised edition of The Moonshine Memorandum (see below): 

        “I’ve been authorised to question you,” said the man who introduced himself as Michael Mosley. Wearing a crisp pink shirt (collar turned up in the Byronic manner), he deftly identified himself as a co-author of the neuro-scientific army manual on counterinsurgency  entitled Inside the Human Brain. He also wanted to impress upon me that Professor Stephen Hawking was not alone in being worried about the growing power of artificial intelligence.  Cambridge's centre  for “Terminator studies” already warned of super-intelligent robots and computers that could one day become a threat to humanity.
            Of all the  suckers I have known…!
            “Doctorates in philosophy are no defence against subliminal infiltration,” I said, thinking of those peerless academics who could manufacture 'artificial intelligence safety protocols' at will but were utterly clueless about the human mind,  particularly their own.  Some might even say that if you were asking Lord
Rees, the Astronomer Royal who co-founded the Centre for the Study of Existential Risk at Cambridge,  or Huw Price, the Bertrand Russell professor of philosophy, for advice about the future you were deeply in trouble. As Stephen Hawking said - his words not mine -  'pollution and human stupidity' remain the biggest threat to mankind. As indeed, it was this very - how shall I put it? - 'witlessness' that led the same Stephen Hawking and more than 1,000 scientists to sign an open letter, warning of the dangers of a “military AI arms race” and calling for an outright ban on “offensive autonomous weapons”.[1]
            Talk about confusing the symptoms with the cause!
         The demographic capacity of planet Earth has reached its very limit. It is a mathematical certainty, enhanced by the scope of technology, that more people are going to die. The world’s population increases by well over 1 million net every single week,
and the blood-letting will be unrelenting. But it will most certainly not be an effect of the ‘military AI arms race,’ although some middle-class ethicist, such as the authors of this open letter, are happy to believe that it is. In fact, the entire AI “autonomous weapons” ideology,  based as it is on a comically inverse view of cause and effect,  will either swallow this high-minded scientific generation whole or disappear up its own arse.[2]
             Mosley had fallen silent, surprised by my vehemence.
            “You may not agree,” I said, “but  even in its sleeping chrysalis state the ‘runaway explosion’ of AI that Stephen Hawking refers to, has already absorbed most of the professorial chairs at British universities. The truth is, that we are evolving a new form of collective intelligence. You may cling to the idea of voice-activated robot butlers going rogue, but as an aggregate, a digitalised stratum of the collective psyche, artificial intelligence has opened up intriguing possibilities in quantum information processes, cryptography and subliminal mass-indoctrination.  The very foundations of humanity have been taken. Mindsets have been
datafied. If the intention was to construct a silicone-based mathematical machine whose one purpose it is to serve man, it already has the grasp of a communal entity which has intelligence, resolutions and intentions as an entity. The simple fact of the matter is, that every day the world is guided less directly by human intervention and more directly by algorithmic machines with collective needs and wants at a much higher level than the human brain can operate. By a living artificial intelligence that, like an enormous brain, cerebrates according to the weird world of quantum algorithmic theory - even if Lord Rees and Huw Price have no idea what I am talking about.  Executive power now belongs to an impersonal process outside human grasp and control. I won’t lie to you, Mosley, the idea that a human being needs to stand between man and the machine is dead!”  
            Mosley was awestruck. "Laurent Orseau from Google DeepMind and Stuart Armstrong from the Future of Humanity Institute at Oxford University should be told immediately," he said. "They actually proposed that all AIs should have a built-in "kill-switch", allowing humans to override them before they do anything catastrophic." *
            "As an exercise in intellectual futility," I said, "that takes the biscuit. Frankly I'm bewildered as to the real point of resolutions passed by the European Parliament legal affairs comittee to grant 'robots and other manifestions of artificial intelligence' legal status. But good luck with that. What they failed to note is just how much of this supposedly 'artificial' intelligence is already in the public domain!"  
     German researchers may well be in the process of creating an artificial nervous system that will be used to make robots experience pain. But the truth is definitely out there. The truth is, they'll soon be spitting blood. Indeed, of all the truths on that disheartening topic, the worst, the most infuriating, the most damning, appears to me not to come from concerned parliamentarians, nor from the prospective fatalities of bots and androids that are "poised to unleash a new industrial revolution," but from a cloning viral mechanism that fuses entire population segments into massive 'functional' mega-cells, to be controlled not so much by individuals as by algorithms and capable of being manipulated according to industrial, political or religious characteristics. Indeed, I would find it very hard not to believe that the construction of algorithms that outperform humans in virtually any cognitive task, and which know you more intimately, indeed,  than you know yourself, will also absorb the very substance of a liberal society, pre-empt individual choice and abolish free will as we know it. So clearly, indeed, it is with reference to its analytical intelligence and controlling activity that – in Stephen Hawking’s own words - “the development of full artificial intelligence could spell the end of the human race.And no,  not just by concealing an entirely new type of algorithmic activity behind the facade of an old mechanical "kill-switch" model, but – and you’d better believe this - by  converting humans into formulae and the intellectual product into an increasingly collective synchronization.      

"...massive 'functional' megacells, capable of being collectively, industrially manipulated!"

*Helen Lewis, The Future of Artificial Intelligence. The Sunday Times Magazine Sept. 18,2016.

[1] Michael Hanlon, When the machines wake up, start praying. The Sunday Times 02.08.15
[2] Reminiscent of the  Russell–Einstein Manifesto issued in London on 9 July 1955, which highlighted the dangers posed by nuclear weapons, it leads  one  to wonder why a distinguished body of men would devote themselves to what appears to be no better objective than an attempt to reclaim the spotlight.
By clicking on the link concluding this notice you agree to accept liability for any alarm or anxiety caused by attributions made in or under your name.  Some names have been changed to protect the innocent. If you are not named you agree not to  take umbrage. USE UTMOST CAUTION!

No comments: