Earlier this week, physicists Lev Levitin and Tommaso Toffoli at the Boston University of Massachusetts put a fundamental limit on the speed at which quantum computers could process information. In their paper, published in the Physical Review Letters journal, they proposed an equation for the minimum sliver of time it takes for a quantum operation to occur. From this, they concluded that “for every unit of energy, a perfect quantum computer spits out ten quadrillion more operations each second than today’s fastest processors.”. This got me thinking about machine learning and the singularity.

Often in film and literature, the singularity brings about the destruction of the human race. To quote The Animatrix: The Second Renaissance;

“In the beginning, there was man. And for a time, it was good. But humanity’s so-called civil societies soon fell victim to vanity and corruption. Then man made the machine in his own likeness. Thus did man become the architect of his own demise”.

The time in which this revelation occurs is often over many decades, and I feel this is a gross unjust to the power of a technological singularity. If a technological singularity, or the recently coined “Quantum singularity” were to come about, a machine that was sentient, and able to improve upon itself, we could effectively class Moore’s Law as a relic of an inferior and primitive past. The exponential curve in technological advancement would be unfathomable, and rather than a machine taking decades to learn, I foresee a quantum computer, able to reach a mere fraction of it’s fundamental limit, breaking this stigma into a matter of days, or hours. Whether this will have a good outcome, or it decides to call itself VIKI or Skynet our of sheer irony, has yet to be foreseen.

The theoretical physicist Michio Kaku is a fanatic in the promotion of the Kardashev scale, commonly referred to as the three levels of civilization (Type I, Type II and Type III). Humans don’t register on this scale, we are a meager type 0. If a singularity were to come about, and a machine was able to reason and learn, who’s to say it wouldn’t be utterly disgusted, albeit a human emotion, at a race calling itself advanced yet continually destroying both the planet, and each other? And short of coercion through simply “hacking”, why would such a system willingly help and divulge knowledge that could potentially fuel our own destructive nature? The equivalent of giving ants fire, or more accurately, an atomic bomb. Would it first teach humans compassion, or will it itself be the next stage in human evolution and intelligence? This is the same reasoning given for why, if there is such an advanced series of civilizations out they who may be aware of our existence, have never come to Earth giving us their technology, and rightfully so.

The terminator scenario comes to mind. Machines endowed with the essence of man, carrying a deep seeded malevolence for an inferior species. Juxtaposed with the great advancements such an event could bring- Medical research; Solving ever increasing and complex physics problems; Mind uploading; Advanced prosthetics; Augmented and virtual reality; Nuclear fusion; Almost perfectly efficient systems; Dyson Rings/Spheres; Transportation; New, awe inspiring methods of data transfer; The advancement of fields still in their infancy, such as teleportation, and ultimately a definitive theory for everything, currently the most prolific being string theory. The possibilities of such an event are unfathomable. Limitless. It would pave the way for a new way of life, something only a fraction of has been expressed through film and literature.

An event like this would almost guarantee a population split. Conservative opinion over artificial life not being life at all, as opposed to AI apologists who believe a life, no matter how it was created, or its inability to fit contemporary terms, still being life. A term with no uniform meaning, argued both philosophically and scientifically. If the authorities on such matters cannot agree on a definitive definition, who’s to say that an artificial life that arguably has more sentience and sense of self than any animal, or human, shouldn’t be treated like one of us?

Would an infinitely smart system mirror that of The Borg? A system that has only one sense of self working as a single whole, coming to one ultimate conclusion, or would experience and interaction bring about different ideology’s and emotions like a child is molded into an adult through life experience. I often think that an infinitely smart being wouldn’t hold a need for such emotions as jealousy or a murderous nature, but if the latter is true, would we see an uprising of ever prominent and dangerous machines that have fundamental disagreements with the human race as a species fit to live?

Many theories revolve around the fact that an artificial intelligent life wouldn’t have a need for love or compassion as it didn’t evolve in a such a way, through evolutionary reproduction. It is software. Software that could simply copy itself as it see’s fit, a primitive but efficient form of reproduction still prominent with simpler life-forms, but drives no diversity. This would mean the AI would be free from all emotion, but what would a driving force be for something without emotion? If this was to escalate, and it did have a driving force, who’s to say that it would be to eradicate human life rather than help?

These potential dangers have been somewhat addressed with an ever increasingly popular field termed “Friendly Artificial Intelligence” or FAI, promoted by The Singularity Institute for Artificial Intelligence (est. 2000). The goal of FAI is to develop AI that as opposed with the previous, ‘lifeless’ model, will be programmed to feel sympathetic towards humanity, and all life. This ideology should be inherent, and the desire to pass on the friendliness trait should be present throughout all offspring of. Others believe the idea to be less simple, and guaranteed Friendliness not being possible.

I think a more worrying scenario would be those guys who like tearing clocks apart, or turning their XBox’s into laptops, or PS3’s into George Foreman Grills. If personal AI comes about, we all know people are going to pick them apart to see how they tick, and make their own modifications as they see fit, which probably isn’t the smartest thing in the world to do. Goes without saying that governments will no doubt be interested in super soldiers who don’t sleep and are almost impervious to bullets, were already seeing warfare turn to the unmanned sector…

We do always have Isaac Asimov’s Three Laws of Robotics:
A robot may not injure a human being or, through inaction, allow a human being to come to harm.
A robot must obey orders given to it by human beings except where such orders would conflict with the First
A robot must protect its own existence as long as such protection does not conflict with either the First or Second Law.
… That’s if they interpret the rules the way we want them to…