The advantages of computer programs over humans

The digital brain has several advantages over the biological one:

Editing capability.  Changes in any parameters in the computer brain occur quickly and easily.

Reproducibility.  Making and saving the required number of copies of a digital document or program is a matter of minutes, and the information stored in the biological brain is reproduced very slowly and inaccurately.

Achievements of goals.  The path to achieving goals in a human team is determined by the individual characteristics of each person and his understanding of goals, while programs are devoid of emotional and cognitive characteristics.

Information exchange.  People get information through learning, and machines just exchange files.

Artificial intelligence emergence rate

The transition to the superintelligence level can be carried out according to one of three scenarios:

Slow development  will take tens or hundreds of years, and people will be able to adequately prepare for the meeting. During this time, it is possible to develop and test different approaches, train specialists and invent mechanisms for protection and interaction.

Rapid development  - the superintelligence will appear suddenly, within minutes, hours or days, and people will not have time to do anything. Already today, an action plan should be developed in case of rapid developments.

Moderate development  - a new superintelligence develops over several months or years, and people have a chance to give a decent answer, despite the tight deadlines that limit the ability to analyze, test different approaches and coordinate actions.

The rate of development of the intellectual abilities of the system is a monotonically increasing function of two variables: the strength of optimization and the response of the system to this optimization:

Intelligence rate of change = optimization strength ÷ resistance.

Supermind Forms

Technology and some animals are superior to humans in many narrow areas.

Bats perceive sonar signals better than humans, calculators calculate faster, and special programs beat people at poker.

The superiority of programs will expand and strengthen, however, we will be able to talk about the emergence of a superintelligence, only if this intelligence is significantly superior to humans in most areas of cognitive activity. The future superintelligence is likely to take one of three forms:

A high  - speed superintelligence is a system that can do the same thing as a person, but much faster.

The simplest example of a high-speed superintelligence is a full-scale brain simulator controlled by powerful hardware. The model operates at a speed ten thousand times the speed of the human brain. She can read a book in a few seconds, write a doctoral dissertation in four hours, and do everything a person does in a lifetime in a couple of days.

It is difficult for a high-speed superintelligence to communicate with people - they simply cannot keep up with it. If communication is necessary, a nanoscale manipulator will come to the rescue - a micro-robot that performs delicate work.

The collective superintelligence  is a system that achieves high results by combining various electronic minds. The collective superintelligence is effective at solving problems that can be broken down into pieces.

Tasks like building a spaceship or setting up a chain of fast food restaurants provide many opportunities for the division of labor: each developer is responsible for specific components of the shuttle, and different restaurants have different chefs, waiters and cleaners.

A high-quality superintelligence  is a system that is smarter and no less fast than the human brain. It is difficult to imagine what a high-quality superintelligence looks like, since a person has not yet had to deal with systems that surpass him in intelligence.

Obviously, a high-quality superintelligence will intellectually surpass humans to a greater extent than humans are superior to elephants, dolphins, and chimpanzees.

A superintelligence existing in one of these forms will be able to independently improve and create various technologies.

Age of supermind

Now it is not easy to imagine how the superintelligence will behave and how it will end. Nevertheless, futurological research allows us to put forward more or less realistic hypotheses about the future of artificial intelligence.

Supermind superpowers

The superintelligence will act faster and more efficiently than the best minds of humanity, thanks to the corresponding superpowers: computers questions

strengthening one's own intelligence  - programming and reprogramming artificial intelligence and research to improve the efficiency of cognitive activity;

strategy development  - achieving goals through forecasting, prioritization, planning and analysis;

social manipulation  - creating social and psychological models and persuading people;

hacking  - detection and exploitation of weak points in the security system of computer systems;

technological research  - modeling and design of nano and biotechnology;

economic efficiency  - the organization of productive intellectual activity.