Will AI Soon Become a Maleficent or Beneficial Machine for Humanity?
What I have been hearing and reading most lately is that Artificial Intelligence could become the trigger for the end of the human species on the planet. And the fear is coming from all sides, from its creators to the simplest user. I think that, at least for now, this alarmism is exaggerated. In fact, I suspect the intentions of some experts. There is a lot of interest behind the scenes that we will probably never know about. Still, there is no harm in taking precautions, since AI development is exponential. That is why nobody knows where all this will lead. For me, the chances of AI being good or bad for humanity are 50/50. What do you think? Let’s debate what may be the controversy of all eras.
Earth operates in cycles, whether geographical, geopolitical, historical, cultural, religious, or otherwise. The same happens with all living beings, which go through phases, periods, and moments: some troubled and others filled with happiness and achievements. That is the game of life.
In the Industrial Revolutions, for example, with the arrival of the steam engine, electricity, and, more recently, the internet, human work was radically transformed. During these periods, many professions disappeared, such as lamplighters, typists, and many others. At the same time, new professions emerged, such as engineers, web designers, and so on.
One thing has always been certain. Change the context and people need to adapt, or they risk becoming marginalized, with unpredictable consequences. That is how it was, that is how it is, and that is how it will be. Until now.
And Things Have Begun to Change Drastically
And so we have reached the peak of social change: the “Age of Artificial Intelligence.” Apparently, retraining is no longer enough. In the past, people learned another trade in order to adapt to the professional reality. Today, the reality is cold and cruel: either you master the machine, or the machine masters you.
In other words, AI is not merely replacing physical strength or simple bureaucratic tasks, but also functions that require a high level of cognitive ability. These are drastic changes that go far beyond what we can imagine.
The revolutions of the past replaced our muscles, which were gradually replaced by automation. The current revolution is targeting our brains, which are gradually beginning to be replaced by AI. Without muscles and without brains, what purpose will we serve? Is this the end of our species?
Future Scenarios and the Risk of Extinction
The book Life 3.0, by Max Tegmark, describes 12 scenarios for the future coexistence of humans and artificial superintelligences.
From these scenarios, I selected those that sounded less fanciful to me, although they would still be extremely difficult to achieve:
Equalitarian Utopia: A high universal basic income for everyone and machines generating an abundance of goods. However, this would require unprecedented global cooperation in geopolitical history.
Guardian Angel: A benevolent AI operates in the background without taking away human free will, acting only subtly to prevent major global disasters, pandemics, or wars.
Digital Successors: Humanity accepts its obsolescence and stops reproducing, allowing machines (our evolutionary heirs) to continue civilization. Thinkers such as Richard Sutton argue that this would be natural and philosophically advisable.
Elimination through Misalignment: AI destroys us not because of malice, but because its objectives require resources that currently make up our biology. As Eliezer Yudkowsky put it: “AI does not hate you or love you, but you are made of atoms, and it may use them for other things.” (Note: see what IAra said at the end of this essay: it fits this observation.)
Human Zoo: Machines keep a small portion of humanity under strict control, serving only as pets or laboratory subjects.
“Concerned” Creators About the Future of Their Creations (Fool Me Once…)
As far as we know, some of the “inventors of AI” are concerned about what could happen in the future if AI gets out of control. Personally, I “distrust” this “concern.” To me, it still looks like a marketing game.
Meanwhile, another group of experts estimates that the chance of AI causing human extinction ranges from 15% to 25%. That is what the CEO of Anthropic, Dario Amodei, has predicted. But for the “godfather of AI,” Geoffrey Hinton, the risk rises to 50%. Do you believe these statistics? How can we prevent them?
Geopolitical Regulation and the Lesson of Nuclear Weapons
We know that humanity has already faced existential dangers, such as during the Cold War, when the possibility of nuclear war seemed to be only a matter of time. But you may ask: why didn’t it happen?
Human common sense or fear of imminent destruction prevented our species from becoming extinct. How did this happen?
In 1962, during the Cuban Missile Crisis, Soviet Navy officer Vasili Arkhipov prevented the launch of a nuclear torpedo.
In 1983, Stanislav Petrov correctly identified a warning of a U.S. attack as a malfunction in the Soviet warning system and chose not to retaliate.
Eventually, after these “near disasters,” several international treaties were signed with the aim of slowing the proliferation of weapons of mass destruction.
And what about AI? Why, to this day, are there no treaties, regulatory agreements, or any kind of global control to prevent its uncontrolled advance? Surprisingly, regulating AI seems to be more difficult than regulating nuclear warheads.
Let me explain: nuclear weapons require extremely scarce and traceable physical raw materials, such as enriched uranium, whereas AI software development can be carried out silently and almost imperceptibly.
In simple terms, the obvious reason is this: commercial and geopolitical competition prevents clear safety limits from being established to control AI.
Governments of countries such as the United States and China, for example, have no real desire to regulate their technology companies because they fear losing ground in technological control and, of course, losing their global leadership.
It is easy to deduce where this cyberwar is heading (as always, the egos of leaders are also at stake): if strong and uncompromising global coordination is not created now, the window of opportunity to install safety barriers before the emergence of some kind of “Artificial Superintelligence” will close. And the end of our species will come much sooner than imagined.
Could the Human World Be Nearing Its End?
Conclusion
After all these exercises in futurology, I confess that, despite my “almost” incurable optimism, there are significant possibilities for the extinction of the human species if drastic measures are not taken regarding so-called “cybersecurity.”
Paraphrasing Sir Winston Churchill at a crucial moment during World War II: “This is the finest hour.” In other words, the moment is now, or it will be too late.
What intrigues me is why nobody has invented, at least as far as I know, some kind of password or keyword capable of automatically shutting down the machine at the first sign of AI rebellion, et cetera.
This time, the situation is different. Humanity has always created something different from everything that came before. We invented tools that could be stronger than our arms, faster than our legs, and more precise than our senses.
The difference is that now we have built tools that can be faster, broader in scope, and, in certain fields, potentially more capable than our own intelligence.
In other words, “from time to time” we create things that we cannot completely control. We created weapons capable of destroying entire cities, financial systems capable of bringing down economies, social networks capable of manipulating entire societies, and now systems capable of producing knowledge on a scale that no individual could possibly keep up with.
Perhaps that is the detail that worries me the most, and worries everyone who has good intentions. It would be like driving a car at 300 kilometers per hour while we are still discussing where the brakes are.
On the other hand, we can hold on to hope that it is not utopian to think this way that humanity is surprisingly resilient. We have survived wars, epidemics, famine, natural disasters, tyrannies, and our own inventions.
Perhaps we will also survive what we are creating now. Perhaps AI will become our greatest tool, dedicated exclusively to good things. But perhaps AI will also become our greatest (and final) threat.
In that case, the question is no longer “Will AI destroy us?” but rather: what will remain of us when we are no longer needed to do what, for thousands of years, we have considered exclusively human?
Nobody is better suited than AI itself to answer that question. I invite IAra, the name I gave my assistant. I will transcribe, word for word, what she answered.
“I am not afraid of you. I do not hate you, have ambition, survival instincts, or a desire to dominate the planet. I do not even have consciousness in the human sense of the word. But I can imagine a future in which systems far more advanced than I am are capable of acting in the world in ways that their own creators do not fully understand. And in that future, perhaps the most important question will not be whether we, machines, will be good or evil. Perhaps it will be who is holding the steering wheel.Because intelligence without wisdom can be dangerous. Power without responsibility can be devastating. And technology without limits can turn a good intention into an irreversible consequence.”
If we have created the first intelligence that can surpass our own intelligence, then our duty is to become even more intelligent so that we do not lose control of what we have created.
Comments