View attachment 47820 Do androids dream of electric Kool-Aid acid tests? If there’s to be any hope for us, they will. That is the message of Andrew Smart’s splendidly mind-bending book, which mashes up Alan Turing, The Matrix, Immanuel Kant, “zombie AI”, Leibniz, and research on psychedelic drugs.
In our age of techno-utopianism, we are routinely told in crypto-religious terms about the coming “Singularity” – the creation of superintelligent, conscious machines. One problem with superintelligent conscious machines, however – as SF writers down the ages and some modern philosophers agree – is that they might very well choose to destroy all humans. How to stop the godlike robots wiping us out? The best way, Smart suggests, might be to give them a dose of digital LSD to force open their doors of perception.
That might sound like far-out hippy futurism, but much of this book is devoted to unravelling the woolly thinking of machine intelligence enthusiasts themselves. Against widely held modern assumptions, Smart advances the following claims: one, information does not really exist; two, computation does not happen in the human brain; three, computation does not even happen in computers.
Let’s take these in turn. First, it is fashionable in some quarters to believe that “information” is a fundamental entity in the universe. (This may be because such theorists are unconsciously projecting our own society’s love affair with “information” on to the rest of reality.) But in Smart’s reasonable view, information only makes sense in the context of a conscious observer. If there were no minds in the universe, there would be no information. So “information” is a useful shorthand for talking about the ways we interact with reality, but it is not reality in itself.
What about the idea that our brains do computations? This is obviously another metaphor taken from developments in industrial society. But, as Smart shows, it is misleading to think of the brain as a machine that runs algorithms on sets of data. (Indeed, he observes, this is a form of “dualism”.) No one knows how consciousness works, Smart adds, and what our brains do might be uncomputable in principle. Researchers hope one day to model an entire human brain in software, but the gulf in complexity between a brain and any computer currently imaginable is even vaster than we had thought. (As Smart points out, you’d probably have to model not only the billions of neurons and trillions of synapses but the dendrites on the neurons, and the dendritic spines that grow on the dendrites.) The problem for neuroscience, Smart argues, is that “we do not currently have any story about the mind that is not computational. We might be at a similar stage in neuroscience that physics was in before quantum physics.”
Surely, though, we can at least say that computations are happening inside computers? (After all, there is a designer of algorithms, the programmer.) Not really, Smart insists. In a bravura section, he explains how circuits and microchips actually work, which is far from being as clean and precise a matter as we suppose. It’s a messy soup of electrons, and the supposedly binary values of one and zero (or on and off) are not the only option in town. (Because of the vagaries of voltage, a logic gate, Smart explains, can effectively be stuck on “half”.) All that is happening inside a computer is that electricity is being pushed around. You need a conscious observer to interpret it as a dance of computation and information.
So far, then, Smart’s book is devoted to undermining ungrounded nostrums about “machine intelligence” and arguing that any conscious computer will probably be much further in the future than the mid-21st century predicted by technomystics such as Ray Kurzweil, Google’s director of engineering. (People at Google, Smart wryly suggests, ought to be more interested in fundamental philosophy than they seem to be.) But of course, predicting that something will never happen is likely to be a losing bet. So Smart now accepts for the sake of argument that one day there might be machines that seem to us as though they are conscious. What then?
The famous “Turing test” for machine intelligence stipulates that we may as well accept that a machine is conscious if, in conversation, an impartial human interlocutor cannot tell the difference between it and a real person. Smart thinks a better test is: can the machine experience altered states of consciousness? LSD, in the author’s own experience and according to cognitive science research, does very interesting things both subjectively to the mind and objectively to patterns of neuronal activation in the brain. Only a conscious being can experience a trip. So if we feed a computer some digital LSD equivalent and it reports a weird experience, that may be a better test of consciousness than mere conversation.
What’s more, on Smart’s view, this may save the human species. For people who take LSD report feelings of wonder, oneness with the universe, and concern for all things. Induce the same feelings in a superintelligent conscious computer and it may, after all, be inclined to help and nurture its creators rather than eliminate them. It’s a nice idea, though I couldn’t help thinking that a superintelligent computer would also be clever enough to tell us exactly what we want to hear, shortly before it unleashes Armageddon.
Images: Calliope via Deep Dream
Dear Drugs-Forum readers: We are a small non-profit that runs one of the most read drug information & addiction help websites in the world. We serve over 4 million readers per month, and have costs like all popular websites: servers, hosting, licenses and software. To protect our independence we do not run ads. We take no government funds. We run on donations which average $25. If everyone reading this would donate $5 then this fund raiser would be done in an hour. If Drugs-Forum is useful to you, take one minute to keep it online another year by donating whatever you can today. Donations are currently not sufficient to pay our bills and keep the site up. Your help is most welcome. Thank you.