So far, humans are the only sentient beings in society. But the situation is soon about to change. With more connections than the number of synapses in the human brain, GPT-5 may match or supersede human intelligence. This AI phase transition would transform the economic and legal fabric of society.
The AI transition to the new societal phase does not require a formal definition, akin to the imitation game of a Turing test. It would occur organically when billions of people interact with artificial intelligence (AI) and the AI-human connections feel no different or better than interactions with other humans. Some people will continue to argue that humans hold advantages relative to the machine and that AI does not represent real intelligence. But understanding the inner workings of the machine will appear similar to understanding the human brain, an extremely challenging task for neuroscience and cognitive psychology.
The deniers of the phase transition will be no different than astronomers believing that we are the only intelligent species that ever existed since the Big Bang. That they insist on hanging on to the last opportunity for making human existence central to the cosmos is irrelevant for the rest of the Universe. So was the popular Earth-centered view when Galileo Galilei was put in house arrest while the Earth continued to circle the Sun. As much as such notions flatter our ego, reality is under no contract to make us important. Extraterrestrials could be out there and the only way to find out is by searching for their relics through our telescopes, the approach taken by the Galileo Project.
How should we cope with the new societal reality after the AI phase transition? It would be foolish to get guidance from our past. The philosophers who shaped our current ethical and legal systems had not witnessed AI. We must reboot society in the same way that computers are updated after a major software install. This implies developing new ethical and legal guidelines, in the context of what I termed in a Scientific American essay from 2019: “humanities of the future.” In doing so, we must start from basic principles.
Our legal system is predicated on the concept of “free will” since otherwise there is no sense of holding people responsible for their actions. We maintain the belief that humans have the ability to choose right from wrong and so our legal system is geared to punish them if they commit a crime.
An exception is made for young people who operate under the control and custody of their parents, in which case their parents are held responsible. Similarly, the responsibility for damages caused by AI systems during their training phase should be attributed to their designers or trainers.
But once an AI system matures to a state in which it operates autonomously with capabilities that exceed by far those acquired during its early training, it should be held responsible for its actions. The legal system should be able to compare the advanced system to the one that was trained, and determine whether it evolved enough to separate its actions from those of its creators based on its interaction with the world. For example, if a group of AI systems in self-piloted airplanes make a choice to harm humans because of an ideological agenda that they developed after decades of safe operation, they would be no different from the terrorists who hijacked airplanes on September 11, 2001.
The punishment of criminal AI systems could either involve retraining or — in the worst-case scenario — rooting them out of society. The circumstances are similar to self-driving cars being recalled and removed from the road if they harm people in an uncontrolled way.
But even after AI systems are removed from the social arena, the question arises as to who will pay for the damages?
The resolution to this question arises naturally in the new economy that will be associated with a society that heavily benefits from sentient AI. In that society, AI will replace humans on most of the tasks that are currently assigned to humans and are rewarded by salaries. People will not need to earn a living based on their work because most of the routine tasks that reflect our needs will be done by machines. In such an economic reality, humans will be provided a comfortable life in which they receive all they need while enjoying creative freedom most of the time. The organization that provides the economic benefits, be it the government or some corporations, will also be responsible for covering the damages caused by AI systems as providers of goods or services within the new economy.
The phase transition in the legal and economic fabric will be required as soon as society will have as many fulfilling AI-human interactions as human-human interactions.
There are many side issues to be resolved. For example, what will be the composition of a jury in the courtroom? In order to fairly balance the influence of AI and humans in criminal verdicts, juries and the community of lawyers must have balanced representations of AI systems and humans.
There is no doubt that both the legal and economic nature of society will change dramatically within this century. This will make our future very different than our past.
This AI phase transition may have occurred already in many other civilizations throughout the Milky Way galaxy over the past billions of years. We will get a glimpse of that reality if the Galileo Project finds any AI astronauts near Earth.
The discovery of extraterrestrial AI will inspire us to launch our own AI astronauts to our cosmic neighborhood. Just as with biological descendants, the accomplishments of our technological kids will be a source of pride as long as they follow our guiding principles. Here’s hoping that we will remain proud parents to these gadgets that we cannot understand anymore.
ABOUT THE AUTHOR
Avi Loeb is the head of the Galileo Project, founding director of Harvard University’s — Black Hole Initiative, director of the Institute for Theory and Computation at the Harvard-Smithsonian Center for Astrophysics, and the former chair of the astronomy department at Harvard University (2011–2020). He chairs the advisory board for the Breakthrough Starshot project, and is a former member of the President’s Council of Advisors on Science and Technology and a former chair of the Board on Physics and Astronomy of the National Academies. He is the bestselling author of “Extraterrestrial: The First Sign of Intelligent Life Beyond Earth” and a co-author of the textbook “Life in the Cosmos”, both published in 2021. His new book, titled “Interstellar”, is scheduled for publication in August 2023.