In 1993, computer scientist and science fiction author Vernor Vinge wrote an essay, “The Coming Technological Singularity”. He predicted, “Within 30 years, we will have the technological means to create superhuman intelligence. Shortly after, the human era will be ended.”
There’s a decade to go and Vinge seems to be on target with respect to technological means. For “superhuman intelligence”, the current state of computing would have to be exceeded by magnitudes. Doing this by increasing processing power and data capacities of conventional computers is difficult.
The most likely route to boost artificial IQ is through quantum computing. The weird effects of quantum physics can be used to develop computers exponentially more powerful than conventional computers. A series of breakthroughs have taken quantum computing from being a science-fiction motif to early proof of concepts.
In classical physics, an electronic circuit is on or off. Conventional computing uses the binary property by assigning “0” (off), or 1 (on) to every computer “bit”. A quantum bit, or Qubit, can, however, be both on and off at the same time, via the principle of superposition, which Schrodinger illustrated with his thought experiment of the cat that is both dead and alive. A Qubit may be either 0, 1, or both, or anywhere in-between. The principles of quantum entanglement mean that the state of one particle influences the state of another. This allows for information teleportation, since two physically separated particles change state at the same time.
A quantum computer with n qubits (where n is any number) can be in a state of superposition of up to 2 to the power of n different states (2n) simultaneously, compared to a normal computer that can only be in one of these 2n states at an instant. Hence, while eight bits (one byte) can store any one of 256 different strings of information, eight Qubits can store 256 different strings at the same time. This allows parallel processing — the solving of many problems simultaneously.
The difference in information storage grows massively in larger systems. A 250 Qubit system could store more information than there are atoms in the universe. The technical problems of handling quantum data are considerable. There are also issues in creating quantum algorithms to manage quantum data, translate it back and forth to conventional classical data, etc.
Some answers have started coming to light and 2012 in particular saw major breakthroughs on different fronts. The physics Nobel was shared by Serge Haroche and David Wineland, two pioneers of quantum experimentation.
Their work demonstrates how superposition and entanglement can be controlled, observed and manipulated in practice. Wineland traps ions (electrically charged particles) inside “cages” of electrical fields and zaps them with lasers. The energy levels of subatomic particles are separated by discrete (quantum) amounts. The laser bombing causes superposition, which forces ions to exist in two energy states simultaneously.
Incidentally, one spinoff is the development of super-accurate ion-clocks.
Haroche has developed methods to trap photons (light particles) and use entanglement to observe photons without destroying them. He bounces photons off two mirrors, which face each other. Haroche induces entanglements between a special type of large atom and photons. Then, by observing the atom, the state of the entangled photon is deduced without destroying it.
In 2012, IBM announced it had fabricated a stable Qubit chip. In IBM’s demo 3 Qubit chip, the Qubit stays stable for about 100 microseconds (one micro second = one millionth of a second). This is long enough to store and query data.
Meanwhile, a collaborative effort between the University of Waterloo, Ontario, and the Vienna Institute for Quantum Optics and Quantum Information achieved long-distance teleportation. It sent data from one Canary island to another one across 143 kms. This shows that the transfer of quantum data via satellites is possible. Teleportation also requires accurate clocks, which is another place where Wineland enters the picture. Canary Islands and other teleportation experiments have mostly been subatomic. In 2012, a Heifei University team demonstrated teleportation of larger objects. They entangled two ensembles, each of which had about 100 million rubidium atoms. The ensembles were physically separated and connected by a 150-metre fibre-optic cable. This implies quantum routers may be developed.
Another useful experiment was performed by Toshiba’s Cambridge Research Lab. In this, encrypted quantum signals were sent on a standard broadband network and deciphered. The commercial implication is that an upgraded version of current telecom networks can handle quantum data.
A new area of research is the boson-sampling computer, which could be an intermediate stage to universal quantum computing. “Boson samplers” only handle specific types of problems but they are also potentially more powerful than conventional computers. They seem to be easier to build than universal quantum computers.
The first obvious utility of quantum computers could be in cryptography. In theory, entangled data is unbreakable. A massively parallel quantum computer could also crack most codes in existence. Other parallel processing problems like genetic research, and climate models, should also be more tractable to quantum computing. If this sort of power is available, people will surely find new ways to use it. Let’s hope however, that Vinge missed the mark with his prediction about the end of the human era!