- Elon Musk said AI is like “summoning the Devil.” But one thing is for sure, we should never underestimate the capability of technological progress.
People over the course of human history have always underestimated future technological advancements. There is this consistent assumption that we are at or close to the end of technological developments. As negative people are about current living conditions, our ancestors, in their wildest dreams never would have imagined the relatively amazing lives we live now. The population globally is more educated than ever. There is less malnutrition. Infant mortality rates keep getting lower and life expectancy longer. Poverty levels are far lower than ever. Vehicle fatality rates keep getting lower despite the fact that cars are getting cheaper. Electricity prices are cheaper than ever especially for clean energy sources. Things in the future will get far better; we’ve just barely hit the tip of the iceberg.
Think about this, once society used to think it would be impossible to go beyond 30 mph (the speed of a horse). Then we moved to not beyond 50 mph, not beyond 100 mph, not beyond the speed of sound (740 mph). We’ve now blown way past the level of sound to levels by 2,000 percent. Continuously for the last 50 years we thought we hit limitations in computing power per price, but we kept on blasting through to higher and higher levels. When Moore proposed his law about exponential increase in computing power the scientific community was skeptical. There used to be a time when generating electricity was 100 times the cost by solar compared to coal. Now in most of the United States it is cheaper to generate electricity by solar compared to coal. Battery prices have been plummeting as industries have been scaling production up allowing for cheaper storage of energy and electric cars.
One thing that’s interesting is that it is apparent looking over 200,000 years of homosapien history that technological advancements are happening at a more and more rapid pace. Most humans think about things in terms of a linear scale. But many exponential growth phases seem linear because we only focus on the short-term. Let us think about how a linear trend works starting from a base “1” and adding “2” for every step…. It would go 1, 3, 5, 7, 9, 11, 13, 15. An exponential trend starting from a base “1” and multiplying “2” for every step It would go 1, 2, 4, 8, 16, 32, 64, 128. In the first 3 steps after “1” for both, there isn’t much difference, but it breaks out with time with the exponential trend ending up 753 percent higher (128 vs 15) after 7 steps. If you extrapolate another 7 steps, then it ends up 7,181 percent higher.
Right now, in many technology companies there is heavy investment in artificial intelligence (AI). What is artificial intelligence? Well it’s hotly debated in the scientific community what we should consider artificial intelligence. There are very narrow and very wide scopes for what artificial intelligence is. But let us accept for this article that artificial intelligence refers to being able to think on the level of at least an average human being. This includes the ability to recognize patterns, learn, comprehend, logically make inferences and invent. The intelligence of human beings comes from our brains which are about a kilogram each (less than 2 percent of our body weight). It’s considered in the scientific community that the human brain is the most complex object we know of in existence. One difference between our brains and computers is that our brains are electro-chemical while computers are electrical.
These are some elements so far that are leading us to the concept that machines can match and even beat human beings. For over two decades, we’ve had supercomputers beat world chess masters. Some of the best performing hedge funds and traders use algorithms designed by pattern and price recognition of markets by supercomputers. We have facial recognition technology on many applications. Right now, tests with self-driving technology show that they are “safer” than normal human drivers. However, these are generally not considered artificial intelligence by the scientific community. It’s primarily coded “logic” that can mimic intelligence in certain specific situations.
One thing that is critical to making artificial intelligence is having human brain capacity in processing speed and memory. Human brains operate at 10^17 calculations per second with 10^12 bytes of memory. Right now, only expensive bulky supercomputers have that level of computing power. But as Moore’s Law continues in exponential computer power it will become more affordable and less bulky. Calculations by current Google’s Director of Engineering and leading technologist Ray Kurzweil in his 2005 book “Singularity is Near” implied that Moore’s Law will be able to go to at least 10^50 calculations per second for the weight of the human brain. Recently, Google has announced it would be capable of making a quantum processor by 2030 that can do 10^300 calculations per second. However, it is important to note, that in addition to having the appropriate computing power we also need the “software” that is capable of “intelligence.”
Reinvent and Redesign
It is believed that once artificial intelligence gets to a certain level, it may be able to reinvent and redesign itself. A machine that can design a smarter more efficient machine can lead to a runaway stage. In a runaway stage where each new generation of machines would keep engineering a next more efficient, faster generation of machines at an increasing pace. Eventually it will lead to a near vertical movement up in the charts on technological development where technological development would appear to be near instantaneous. This point is referred to as singularity. Development pace will tend towards infinity though never absolutely reach it. Over the recent years, the concept of singularity has gained acceptance among the scientific community that it will happen. However, there is a large disagreement over when it will happen. For some people it can take place as soon as a few years while others believe in a century, the average year among technologists for singularity seems to be the middle of the 21st century.
How can vastly superior computing power and artificial intelligence help us? Well for one we can have neural computer connections that vastly increase human intelligence. By vastly increase, think of the average person being able to instantly learn all the educational material ever at all universities within a fraction of second. Imagine being able to accurately simulate quantum and subatomic relations of atoms. This would allow us to design enzymes and catalysts that can vastly increase the rate of reactions. Such molecules can be used to take carbon out of the atmosphere and fix it into far stronger than titanium construction/transportation/consumer materials, food, make batteries, and more. We would be able to design far more efficient solar panels. We can produce room temperature superconductors. We would be able to design food such that it tastes as good as we want for taste buds and be as healthy as we want. We would be able to freeze and reverse aging such that we have eternal youth. Aging and death are biological mechanisms for all species to create genetic diversity and can be conquered with singularity. We can have full immersion virtual reality.
The following article is just intended as an introduction to singularity. The implications are far reaching and the technologies capable are probably well beyond what I or leading technologists can conceive of us. Some technologists have noted that uncontrolled growth in AI can be dangerous. The AI can revolt again us. This how it is in the movie “I, Robot.” Leading tech entrepreneur Elon Musk has mentioned before that AI is like “summoning the Devil.” But one thing is for sure, we should never underestimate the capability of technological progress. It is constantly moving ahead forward as it has been for the last 200,000 years of human history at an increasing pace. Some technologists note we may even be able to travel faster than light with “exotic” technologies in the future. To some of you who are older reading this article, would you have imagined that we would today have such a convenient lifestyle with laptops, desktops, tablets, cell phones, GPS, search engines, social networks, ordering online food, Amazon 1-day delivery and more? There are times when technological progression temporarily plateaus and then sees sudden rapid progression from a major (or series of) inventions.
Rahul Ghosh is a futurist and an aspiring sci-fi writer. He graduated from UCLA at the age of 17 but opted not to go to any graduate school. He delved into the world of trading currencies, securities and commodities as he likes to play with numbers.