Escudo de la República de Colombia Escudo de la República de Colombia
UN Periódico Digital

Resultados de Búsqueda:

UN Periódico Digital
Artificial intelligence, detonator of the Fourth Industrial Revolution

In the history of humanity there have been great social transformations induced by the conditions of the world in each specific moment. In some cases, science and technology have prompted changes to society paradigms, one of which was the computer, which inspired the idea of imitating human beings. In fact, at the beginning, they tried to copy the manner in which human beings carried out mathematical operations, and since then it surpassed it by far.

One of the greatest strengths of computers is the capability of performing logical operations that allow them to take decisions in face of diverse circumstances. For instance, the simple human procedure of hailing a taxi on the street or deciding to walk or take a bus. Logic and mathematical logic is the essential factor used by computers which have allowed developing complex decision taking systems, including expert systems and diffuse logic.

The imitation game

AI has evolved thanks to trying to imitate human beings. In his 1950’s book Computing Machinery and Intelligence, Alan Turing –recognized for being the father of computers–proposed a test to evaluate the intelligence of machines. The test consisted of the capability of the machine to show behavior similar to a person and not be distinguished as being a machine in a normal natural conversation. If the tester does not effectively distinguish between a human being and a machine, the machine passes the test. What is tested is not the capability to imitate human beings but the capability to provide answers in a normal conversation.

Between the Second World War and the 80s, great scientific and technological progress was achieved in the area of AI, including pattern recognition, such as acknowledging traits or features in an image or audio through neuronal networks.

In fact, since the 80s there has been a boom in AI systems due to greater research en implementation of systems inspired in natural physical and biological systems, such as:

  • Artificial neuronal networks inspire in the human nervous system
  • Evolutionary IT inspired by Charles Darwin’s evolution theory
  • Collective intelligence systems, also known as swarm intelligence, which emulates social biological beings behavior, such as ants, birds, fish or microorganisms

Historical milestones of AI

Some of the most important historical facts in the advancement of AI have been:

  • A Deep Blue supercomputer program, in 1996, capable of beating World Champion Chess player Gary Kasparov,
  • Marketing in 2002 of the first home robot, manufactured by iRobot, which as of February 2014 had sold more than 10 million units worldwide
  • An IBM robot named Watson capable of replying to answers in natural language. The robot has an extensive database and was built to compete in the famous TV program Jeopardy, where it defeated two human opponents with record earnings.

In 2013, IBM released the first commercial application of Watson in the medical field in support of lung cancer treatment for the Memorial Sloan Kettering Cancer Center in New York.

  • Eugene Goostman, resembles a Ukrainian 13-year old boy, is a chatterbot, developed in Saint Petersburg in 2001 by a group of three programmers. In a 2014 contest, on the 60th anniversary of Turing’s death, a third part of the judges thought Eugene was a human being, obviously passing the Turing test.

Types of artificial intelligence

In general, AI may be classified into three categories:

  • Narrow or weak AI used massively in real work applications such as face recognition used in Facebook, personal assistants, etc. This type of AI is limited to specific tasks.
  • Strong AI, where machines will have the capability to think, at least at the level of one person. Although it is still challenging to achieve this kind of AI as it is difficult for a machine to recognize certain objects or people or identify sarcasm or link words to particular contexts in a conversation. AI strong machines could cause a great revolution as they could have the capability to learn and update autonomously and ultimately gain a level of intelligence superior to humans. This would cause what is known as “singularity”, giving birth to the third type of AI.
  • Super AI, which is any AI that surpasses human intelligence and rapidly self-improve.

Super AI creates fear and also raises the interest of humans. The fear comes if machines conspire and risk human existence by manipulating or restricting access to essential information or services.

Machine learning

Machine learning is a part of AI which looks for methods for computers to learn, particularly creating programs capable of generalizing behaviors from information provided in examples, represented in computer system data.

One of the greatest achievements was back in 2008 when Google implemented a voice recognition application for cell phones and now applied to assistants such as Siri (Apple), Assistant (Google), Alexa (Amazon) and Cortana (Microsoft).

Another application is in driver-less automobiles. Last year Audi announced that its new A8 model would be autonomous using Audi IA and be the first production vehicle to achieve level 3 autonomous driving. Also in 2017, Waymo announced that it had begun tests with autonomous vehicles.

The impact of AI and associated technologies

To face the challenges of the Fourth Industrial Revolution it is important for governments and particularly educational institutions to be prepared to secure training processes for citizens. In fact, educational institutions should begin transforming their curricular programs to respond to a new society and new requirements of the workforce, creating new careers, as data scientists and other professions currently not offered.

Besides the scientific and technological aspects, there are other ethical considerations that are emerging from this new technology. Similarly, as with the Industrial Revolution (1760-1840), this new revolution predicts loss of numerous employment positions, such as in monitoring systems, processes control, customer service, etc.

Although there have been great achievements in language recognition there are still issues in achieving a capability comparable to humans. It has come to a point where achieving more progress is extremely difficult and slow. Also, the capability of machines to perform automatic reasoning tasks is very limited.

Some of the great transformations are the communication, interaction and joint effort of machines (multi-agent systems) with humans. Machines could be able to improve by themselves, obtain materials for their own manufacturing and inclusively build other machines.

If machines continue to provide services to humans, according to industry experts, we would enter into an era of abundance where practically work for money would come to an end. This could be a challenge for human beings not feeling useful and fall into existential crises as they would have nothing to do. Therefore scientists need to develop artificial emotional intelligence to help people reach levels of happiness and emotional wellness.

AI systems lack feelings and ethics, as they are incapable of no distinguishing good from bad and have undesirable behaviors for human beings. There is a true risk, and for this reason, last year a group of experts in Barcelona wrote a manifesto in favor of an ethical and appropriate use of AI.

 

Consejo Editorial