Technological Singularity — Will machines reach it?
“Any sufficiently advanced technology is indistinguishable from magic”. — Arthur C. Clarke(Science fiction writer)
We program machines. Our programs and machine hardware become more advanced, then again more advanced and again more advanced. From two bits: 0 and 1 and transistors, the digital world grown from mere digital calculators to machine learning and artificial intelligence. More and more objects from the physical world are being cloned to the digital world.
The algorithms become so intelligent and so the machines that contain them. The intelligent machines start to create new improved machines. This creates a cycle where the intelligent machines create new intelligent machines. An intelligent explosion occurs. The machines become more and more intelligent at each cycle and they themselves creates the improved hardware to support the improved algorithms. The machines become so intelligent that future thereafter will be highly unpredictable. They would explore the spectrum of intelligence that we could ever imagine. The world, politics, economy, culture and everything changes abruptly and unpredictably. The future where all our speculations will break down. Technological Singularity has been reached.
In physics, there is a point in spacetime where the mass and gravity of a celestial body becomes infinite and warps the spacetime infinitely. This is what happens in black holes. The laws of physics breaks here. Reality reaches a whole new level. This is known as the gravitational singularity.
Similar to gravitational singularity baffling us about reality, technological singularity baffles us about intelligence. The intelligence reaches to a whole new level. The whole Terminator franchise revolves around this theme where machines even build time machines and alter different timelines to reach their goal. Whether it will be for better or worse is another subject. This article will make you think about whether such a singularity is possible.
Thinking about current technological improvements, from the invention of the wheels to that of the spacecraft, from fire to that of missiles, from medicinal herbs to that of keyhole surgeries and genetic engineering and from basic symbolic representations to the algorithms that control the digital world, the whole journey took less than ten thousand years.
For a person who lived ten thousand years ago aircraft may have just been a dream and would not have thought it would be possible in any sense. Our questioning of what is possible is greatly questioned in this case. So what may happen when machines reach a stage where they produce the similar improvements in smaller amount of time, maybe within a decade or maybe within a year. Imagine the impact and think about the wild and unpredictable possibilities. If technology could improve this fast, marvelous and radical changes in technology can be expected. A future which may not be even in the wildest of dreams.
Can there be an intelligence explosion?
If we continue to develop intelligence in machines, there is very much possibility that there will be an intelligent explosion. The only question is how fast it will be reached?
Moore’s law states that the number of transistors in a chip is doubled every two years and this prediction keeps on coming true. This exponential advancement in hardware supports the even more exponential growth in software. When both go hand in hand exponentially there is no good reason to not believe there will be an intelligence explosion.
But would not Moore’s law fail at some point? Would not there will be a point where further number of transistors cannot be confined to same space? Would not the laws of physics prevent intelligent explosion?
It’s true that Moore’s law will fail at a point in time but not in near future. Even if more transistors cannot be introduced to the same space, there will be new ways emerging to go around this. For instance quantum computers will emerge and evolve. What happens when the fossil fuels finish? We will find energy from other resources. More Teslas would come to the market.
So the laws of physics or the lack of enough hardware may not become a good reason that prevents intelligence explosion.
Then there is another point of view that if machines got intelligent, they would do almost all the routine works that are done by humans and much faster. So wouldn’t the economy crash and people become short of revenue which will lead to decreased funding in AI and thus the advancement stops? People feared when computers where introduced. They though most people lose jobs and perhaps some do. But more jobs were also created and the trends became different. Same is the case here, the whole economy or the forces driving it, everything will be changed and a new trend will be set. Maybe humans themselves try to connect with technologies on a biological level in order to achieve transhumanism and there will be no machines or humans thereafter but a new species merged with the biology of neurons and technology of circuits capable of amazing intellect and physiology.
Finally, the speed at which signals pass through a circuit is very much faster than the neurons. The neurons can pass signals at 80 m/s at most¹ but in electric circuits the signals pass at the speed of light. That is a stunning advantage in case of speed and intelligent machines using these circuits may develop faster than we ever imagined. In case of machines space required for hardware is declining and the performance is increasing at the same time. Stunning computational performance can be exhibited within a small space of hardware. When chips get much denser and powerful than our brains isn’t there every possible chance of reaching a singularity? If that is so they will arise a serious question among us: What does it actually mean to be a human?
References
Gain Access to Expert View — Subscribe to DDI Intel
