Superintelligent AI will be impossible to control. Free Technical Library News

NEWS OF SCIENCE AND TECHNOLOGY, NOVELTY IN ELECTRONICS
Free technical library / timeline

It will be impossible to control superintelligent AI

29.09.2022

The idea of overthrowing humanity with artificial intelligence has been discussed for decades, and in 2021 scientists delivered their verdict on whether high-level computer superintelligence can be controlled.

The catch is that in order to control a superintelligence that is far beyond human understanding, it will be necessary to simulate this superintelligence, which can be analyzed and controlled. But if people are not able to understand this, then it is impossible to create such a simulation.

Rules like "don't hurt people" can't be set if people don't understand what scenarios AI has to offer. Once a computer system is running at a level beyond the capabilities of our programmers, then no more limits can be set.

The superintelligence presents a fundamentally different problem than those usually studied under the slogan of "robot ethics". This is due to the fact that the superintelligence is multifaceted and, therefore, potentially able to mobilize a variety of resources to achieve goals that are potentially incomprehensible to humans, not to mention so that they can be controlled.

Part of the team's reasoning came from the halting problem posed by Alan Turing in 1936. The problem is to know if the computer program will come up with an output and an answer (so it will stop) or just loop around trying to find it.

Scholars noted, as Turing proved with clever math, while we may know this for some specific programs, it is logically impossible to find a way that would allow us to know this for every potential program that could ever be written. This brings us back to AI, which, in a superintelligent state, could simultaneously hold in its memory all possible computer programs.

According to the scientists, any program written to prevent AI from harming humans and destroying the world, for example, can decide (and stop) whether or not it is mathematically impossible to be absolutely sure anyway, which means that its can't be contained. The scientists said the alternative to teaching the AI some ethics and telling it not to destroy the world - something no algorithm can be absolutely sure of - is to limit the capabilities of the superintelligence.

“The study also rejected this idea, suggesting that it would limit the possibilities of AI. The argument is that if we are not going to use it to solve problems beyond human capabilities, then why create it at all? If we are going to move forward with AI "we may not even know when a superintelligence beyond our control will appear, such is its incomprehensibility. This means that we need to start asking serious questions about which direction we are moving in," the scientists noted.

<< Back: Curcumin Prevents Vision Loss 29.09.2022

>> Forward: Yeast transplanted with human genes 28.09.2022

Latest news of science and technology, new electronics:

A New Way to Control and Manipulate Optical Signals 05.05.2024

The modern world of science and technology is developing rapidly, and every day new methods and technologies appear that open up new prospects for us in various fields. One such innovation is the development by German scientists of a new way to control optical signals, which could lead to significant progress in the field of photonics. Recent research has allowed German scientists to create a tunable waveplate inside a fused silica waveguide. This method, based on the use of a liquid crystal layer, allows one to effectively change the polarization of light passing through a waveguide. This technological breakthrough opens up new prospects for the development of compact and efficient photonic devices capable of processing large volumes of data. The electro-optical control of polarization provided by the new method could provide the basis for a new class of integrated photonic devices. This opens up great opportunities for ... >>

Primium Seneca keyboard 05.05.2024

Keyboards are an integral part of our daily computer work. However, one of the main problems that users face is noise, especially in the case of premium models. But with the new Seneca keyboard from Norbauer & Co, that may change. Seneca is not just a keyboard, it is the result of five years of development work to create the ideal device. Every aspect of this keyboard, from acoustic properties to mechanical characteristics, has been carefully considered and balanced. One of the key features of Seneca is its silent stabilizers, which solve the noise problem common to many keyboards. In addition, the keyboard supports various key widths, making it convenient for any user. Although Seneca is not yet available for purchase, it is scheduled for release in late summer. Norbauer & Co's Seneca represents new standards in keyboard design. Her ... >>

The world's tallest astronomical observatory opened 04.05.2024

Exploring space and its mysteries is a task that attracts the attention of astronomers from all over the world. In the fresh air of the high mountains, far from city light pollution, the stars and planets reveal their secrets with greater clarity. A new page is opening in the history of astronomy with the opening of the world's highest astronomical observatory - the Atacama Observatory of the University of Tokyo. The Atacama Observatory, located at an altitude of 5640 meters above sea level, opens up new opportunities for astronomers in the study of space. This site has become the highest location for a ground-based telescope, providing researchers with a unique tool for studying infrared waves in the Universe. Although the high altitude location provides clearer skies and less interference from the atmosphere, building an observatory on a high mountain poses enormous difficulties and challenges. However, despite the difficulties, the new observatory opens up broad research prospects for astronomers. ... >>

Controlling objects using air currents 04.05.2024

The development of robotics continues to open up new prospects for us in the field of automation and control of various objects. Recently, Finnish scientists presented an innovative approach to controlling humanoid robots using air currents. This method promises to revolutionize the way objects are manipulated and open new horizons in the field of robotics. The idea of controlling objects using air currents is not new, but until recently, implementing such concepts remained a challenge. Finnish researchers have developed an innovative method that allows robots to manipulate objects using special air jets as "air fingers". The air flow control algorithm, developed by a team of specialists, is based on a thorough study of the movement of objects in the air flow. The air jet control system, carried out using special motors, allows you to direct objects without resorting to physical ... >>

Purebred dogs get sick no more often than purebred dogs 03.05.2024

Caring for the health of our pets is an important aspect of the life of every dog owner. However, there is a common assumption that purebred dogs are more susceptible to diseases compared to mixed dogs. New research led by researchers at the Texas School of Veterinary Medicine and Biomedical Sciences brings new perspective to this question. A study conducted by the Dog Aging Project (DAP) of more than 27 companion dogs found that purebred and mixed dogs were generally equally likely to experience various diseases. Although some breeds may be more susceptible to certain diseases, the overall diagnosis rate is virtually the same between both groups. The Dog Aging Project's chief veterinarian, Dr. Keith Creevy, notes that there are several well-known diseases that are more common in certain breeds of dogs, which supports the notion that purebred dogs are more susceptible to disease. ... >>

Random news from the Archive

Mechanical guards to protect against attacks by wild animals 04.10.2023

Japan has introduced unusual measures to ensure the safety of elderly citizens from bear attacks - mechanical wolves, known as Monster Wolf, have come into action.

Originally created to terrify wild animals on farmland, these terrifying robotic wolves are now being used to prevent bears from attacking people.

According to the manufacturer Monster Wolf, the first use of these mechanical wolves was recorded in Japan in the fall of 2020 after a number of bear attacks on elderly residents. In the latest phase, local authorities across the country are actively purchasing these robots due to the growing threat of bear attacks.

This unusual initiative is due to a massive exodus of young people from rural areas, leaving older citizens without adequate protection and leading to the devastation of cities.

Tokyo University of Agriculture and Technology professor Shinsuke Koike points out that areas that previously served as buffers between bears and people are disappearing, allowing bears to move closer to populated areas and making them less timid around people.

In the Hokkaido region, where brown bears live, more than 150 bear attacks have been recorded over the past 60 years. There have been four fatal attacks and more than 2021 injuries in 10 alone, making it one of the region's deadliest years.

Despite the strangeness of using robotic oxen, they have proven to be quite effective.

See full Archive of science and technology news, new electronics

All languages of this page

Home page | Library | Articles | Website map | Site Reviews