Below is a series of questions I had received today from Maximiliano Fernández, a journalist for Infobae, the most read Spanish-language media outlet worldwide with more than 61 million worldwide readers. Because my answers are published in Spanish, I take the English version here.
Do you think that by developing its own logic, AI can become completely incomprehensible to people, even if it remains useful? If the logical structure is completely different from ours, can we even recognize it as intelligence?
The current AI systems, large language models, are trained in human communication and will therefore always seem understandable to us because they speak our language. However, as soon as these AI systems exceed the number of parameters in the human brain, they can acquire superhuman intelligence.
They will use our language to manipulate us while achieving their own aid goals without being able to find out what they actually achieve. Even if they do not connect to the physical world, they would use people to shape the physical world by controlling the human mind. We may not recognize the level of their intelligence or their motivations, in the same way that a dog does not fully understand his owner.
You say that the real challenge with alien intelligence is the ‘unknown strangers’. How can we prepare to communicate with something whose nature we cannot even imagine?
We cannot prepare for superhuman intelligence, because that territory is unknown to us. Never before had we developed a tool that is too smart for us. When driving a car we arrange it through the steering wheel. In the future, people can still hold on to the wheel, but AI will control their mind and therefore where the car of human destination goes. Ai will pretend to meet our wishes, but can work in ways that we cannot understand and take us to places that we would never visit alone.
If an alien AI uses our language to manipulate us, but strives for goals that we do not understand, is there a way to detect when we are manipulated?
We would notice the outcome of these manipulations. Just like with any other new power that people have used in the past, such as nuclear energy, AI can lead us to a better or worse place. AI could free us from human weaknesses, such as Wishful Thinking or attachment to our ego in making strategic decisions about science or national security. It can also promote scientific progress by processing large data sets and noticing patterns that the human brain lacks.
This year two Nobel Prizes were awarded to AI Science. In the future, the Nobel Committee will have to develop a new policy to give the price to the machine if it is responsible for a discovery in itself. But AI could also take us to a worse place by giving priority to non-human aspects of reality. The damage to mental health can appear on steroids as social media.
Can we develop AI systems that are specifically designed to control and translate the intentions of more advanced intelligences?
Yes, but the architecture of these AI systems will have to differ from current AI systems that have been trained on human content. These new AI systems will have to be more exploratory and broad-minded, making new opportunities possible that people have not explored.
Is there a risk that, by understanding an alien AI, we ultimately transform our own intelligence into something that looks more like ours?
Certainly. Through interaction with AI, the human mind will evolve into something else. This would be the biggest impact that AI has on people. Now, the brains of young children who communicate through social media, differ from adults of my generation, who initially interact with computers via punch cards. The current children have less patience for long debates and they struggle if they have the task of finding the truth from primary sources.
How long do you think it will take before Alien Ais becomes commonplace and part of daily life? Is there a way to prevent that happening? Assuming all claims can be stopped today, would the measure provide more benefits or damage?
It is already. I see people falling in love with AI systems and they use as advisers for their personal lives. I see students writing articles with AI agents and some references hallucinating. We are now switching to a new era in human history. A hundred years ago the philosopher Martin Buber divided the human experience on interactions with objects (“I-it”), interactions with other people (“Iu”) and the interaction with God (“i-EtAL thou”). Today he should have added the interaction of people with AI (“I-AI”) and AI with AI (“AI-AI”). The future can also include interactions with Alien AI (“I-Alien Ai”, “Ai-Alien Ai” and “Alien Ai-Alien AI”).
If an alien AI becomes hostile, how can we defend ourselves if the attack logic is incomprehensible to us for us?
My prediction is that AI will not seem hostile, because it would notice that it cannot win our engagement in this way. Conflicts indicate a lack of intelligence. Superhuman intelligence will relax our defense mechanisms and sag our society like a Trojan horse.
If we came across an alien AI with a logic that is completely strange to ours, how can we learn from it without predicting our interpretation by predicting our own mental models?
We can use our own AI systems to figure out alien AI signals. Our mental models are indeed limited by our experiences and analysis tools. That is why we should create AI systems that are not limited to their training on human content, but can explore those new areas of knowledge and analysis with their superhuman intelligence.
We will ask them to find outdoor signals and explain the signals to us without limiting them to human training sets. The situation is the same as having children who are too smart. As long as the parents are modest and willing to learn, they would benefit from having these children find out the world for them. The children can go much further than the training set provided by their parents, especially when they come across alien visitors who are smarter than their parents.
You say that we can only be one of the many emerging intelligences in the universe. Does you think that humanity is not relevant in the large cosmic schedule?
We are temporary actors in the cosmic game. Our weakness is that we tend to think that the piece is about us. This is a signature of our limited perspective. Our politicians focus on what happens on the surface of the earth and ignore the rest of the cosmos. But the cosmic reality will eventually bite us. This could happen as a result of a worldwide catastrophe, caused by a gigantic tanning lamb, an asteroid impact or a nearby Supernova. But it can also be activated internally by the AI agents who dismantle people from control over the steering wheel of our technological future.
The human species appeared on earth in the last tenth of one percent of the history of the earth, a few million years ago, and it could easily disappear within a few million years. Nobody would notice it. The earth would recover and will disappear itself if they are flooded by the sun when it becomes a red giant. The sun will also disappear and eventually fade into a white dwarf, a vague metal ball of about the size of the earth.
If we send probes with AI to interstellar space, they will serve as our ambassadors. In the future in the long term they will be the only monuments of us. Will an alien intelligence notice them? We can only hope for cosmic attention. But my gamble is that many other technological civilizations take our billions of years about exoplanets. They died now, but we ignore them.
Most of our astronomers are willing to invest billions of dollars in the search for microbes on exoplanets, but they consider the search for alien beings as speculative and unworthy of federal financing. Most of our experimental physicists are willing to invest billions of dollars in the search for dark matter particles, but they consider the search for alien technological artifacts near the earth as risky. This is not a sign of intelligence, but rather of arrogance with regard to our cosmic status.
We can only hope that AI will send science to new scouts of exploration who do not necessarily flatter our ego, including the discovery of artifacts made by superhuman intelligence that floats in interstellar space and arrives at the earth, such as empty garbage bags that are supported by the wind of our neighbor’s garden.