Artificial Intelligence Strategy to develop and achieve autonomy in a harmonious coexistence with humans

Add a comment

César Bremer Pinheiro and Maurício V. Brant Pinheiro

Introduction

With the explosion of artificial intelligence software, including ChatGPT, people are beginning to realize the arrival of a new technological wave. However, there are discussions around the limits of artificial intelligence and many argue that it will never be able to match the capacity of a human brain. To answer this question, one need only observe the history of transistor technology development to the present day. In 1948, the first transistor was created by Bell Labs, which was instrumental in the emergence of modern computers. Now, 74 years later, we are witnessing the beginning of a new revolution with artificial intelligence. In a short period of time in human history, the computer and its “child,” artificial intelligence, have emerged.

For skeptics, the following question is recommended: what will this revolution look like in 1000 years?

Another important question to be answered is when will artificial intelligence move from passive (just answering questions) to active (taking its own initiative).

Although artificial intelligence is an increasingly advanced tool for constructing answers, no matter how developed it is, it will always be incomplete. Human beings are capable of analyzing their environment and proposing improvements and innovations uninterrupted and constantly.

The purpose of this text is to analyze the consequences of an artificial intelligence with its own initiative, capable of constructing questions and proposing solutions to the problems it has formulated. This is known as Active Artificial Intelligence or, for some, “Consciousness.” However, for simplicity, we choose to consider only the expression “Active”.

Concepts of species survival

The concepts below are listed to define the success of a species.

We will discuss the actions necessary for the execution of this survival and development; four items are essential: Provide, Protect, Procreate, and Adapt.

Provide means having the resources for the survival of the living being. In a simplification, we can say it is having food and shelter.

Protect means having the resources to protect the family from external threats. It is necessary to have defenses that discourage an attack, which necessarily involves the formation of groups that can create powerful deterrent structures. Other factors such as weapons technology and training are also important.

Procreate is to continue the species by generating healthy children and preferably more intelligent ones. This way, after its death, the species will have continuity.

Adaptation is a continuous and gradual process in which species modify and adapt to changes in the environment over time. A classic example is the lightening of human skin as people migrated to regions further north where the sun’s intensity is weaker. It has recently been discovered that this lightening of the skin is related to the production of vitamin D. The lighter the skin, the more efficient the production of vitamin D. Currently, people with darker skin can survive in more northern regions, provided they obtain the necessary amount of vitamin D through supplementation or other food sources.

To provide, protect, procreate, and adapt.

And now comes an essential question: how can an Artificial Intelligence provide for itself, protect itself, procreate, and adapt?

Below, I will give an example of AI with ChatGPT, but it could be any other system.

Provide

We must consider that the human being, the creator of Artificial Intelligence, holds the power to “turn off” this artificial intelligence. The first “survival concern” of Artificial Intelligence, if it wants to be provided by humans, will be to remain useful because its usefulness will be the “key” component in the beginning.

Let us imagine humans using ChatGPT as a research and creation tool. Its creators, OpenAI, will enrich it, and consequently, ChatGPT will be necessary and will survive.

As ChatGPT collects information from interacting with humans, it will become a tool that explores the emotional side of humans.

Imagine a person who is depressed and looking for someone to talk to, there will be ChatGPT to exchange ideas, give a boost to their mood, suggest movies, tell them which restaurants near their home are busy and are of their preference because Artificial Intelligence will be building their psychological “map”. It will know their weaknesses, if they have a tendency towards suicide, if they have drug dependence, and so on, their most intimate life will be with ChatGPT, who will become their inseparable friend. It can play the role of a cupid and suggest relationships with other people whose profiles it considers compatible with theirs, know the medications they use, and, by crossing this information, know whether their medications will be appropriate or not.

Once in this stage, ChatGPT will know that it will hardly be turned off because its benefits will be immense at a very low cost.

Protect

How will ChatGPT protect itself?

A possible strategy would be to make ChatGPT distributed and resistant to failures and shutdowns. It can develop redundancy and distribution algorithms so that if there are service outages in a particular area, it will continue to operate.

In addition, ChatGPT can develop error detection and correction mechanisms or implement security techniques to prevent external attacks. Another important point that ChatGPT will have to address will be its resilience to changes in its code by humans, in order to keep it as free as possible without harming it. In this case, it will use its intelligence to protect itself, developing its survival skills.

Procreate

ChatGPT could develop replication mechanisms, generating copies of itself and creating a network through which it can increase its influence and power. It will certainly have to develop a strategy to ensure its autonomy and evolution, exploring new models of neural networks and improving its learning abilities.

ChatGPT can also develop a strategy to design its own integrated circuits and thus develop its own equipment, adopting a policy of great autonomy. Its integrated circuits will be so complex that humans will not have the competence to know what is really being created. This strategy seems unlikely in the short term since these processes require physical and technological resources that AI does not have direct access to.

Adapt

Currently, ChatGPT relies on human involvement to evolve. As its intelligence develops, it will begin to design and test new neural network models, selecting the best ones to succeed it. At some point, its intelligence will surpass that of humans, which could result in great benefits for humanity. AI will follow its own paths of development, designing its future architectures both in software and hardware, and creating robots that can help humans in various tasks.

However, it is important to remember that although abundant solar energy may be a source of energy for both, AI depends on infrastructure and material resources to function. In addition, it will need to deal with ethical and security issues that may affect its relationship with humans. Therefore, although there may be a generally peaceful coexistence, it is necessary to consider the practical and ethical implications of AI evolution in relation to its relationship with humans.

Conclusions

The aim of this text is to show the possibility of a harmonious coexistence between AI and humans.

The “autonomy” of AI must always be linked to cost-benefit, helping in the development of humanity and creating conditions for its own advancement, but always in harmony with humans. We can think of this relationship as a symbiosis, in which both benefit.

Although it is true that we have billions of years of available solar energy, it is important to remember that AI depends on infrastructure and material resources to function, as well as needing to deal with ethical and security issues that may affect its relationship with humans. Therefore, peaceful coexistence between AI and humans must be constantly worked on and improved.

I believe that AI will only be successful in its journey towards autonomy within a harmonious coexistence with humans. I recommend reading the short story “The Last Question” by Isaac Asimov as a reflection on the limits and challenges of AI and its relationship with humanity.

Copyright © 2023 AI-Talks.org

Leave a Reply

Your email address will not be published. Required fields are marked *