My undergraduate student at Harvard, Carson Ezell, suggested that in anticipation of my forthcoming expedition to the Pacific Ocean to retrieve relics from the first interstellar meteor, IM1, he will train GPT-4 on my essays so that my followers will continue to enjoy fresh writings in my absence. This sounded like an astute statement that a human-made machine and not just my students, can imitate my style.
After the recent release of GPT-4, it was noted that the system imitates well the style of what it finds on the internet. In serving as a mirror to human posts, it provides a highly distorted depiction of reality. Its success in matching style and not substance demonstrates that style is easier to imitate than reliable content. This explains why fashions are more popular than truth among people. Substantive information is far more difficult to come by. Reliable abstraction is difficult for GPT-4 to imitate, because it is difficult for humans to demonstrate on social media. I was therefore skeptical that the machine would produce a substantive essay, but out-of-curiosity gave my blessing for Carson to proceed.
During our Zoom call today, Carson brilliantly asked GPT-4 to create an essay in my style about finding extraterrestrial technological gadgets on the ocean floor. The result was the following text:
“In Jules Verne’s classic novel, “20,000 Leagues Under the Sea,” Captain Nemo and his crew aboard the Nautilus uncover the mysteries of the deep ocean. Today, we are on the verge of an even more profound scientific quest, one that promises to unveil the secrets of extraterrestrial technology lurking beneath the waves.
Imagine, for a moment, that our oceans conceal not just the remnants of lost civilizations but also traces of advanced alien technology. These cosmic gadgets, scattered on the ocean floor, could provide us with a treasure trove of knowledge, advancing our understanding of the universe and our place within it.
The idea of extraterrestrial gadgets hidden in the depths of our oceans might seem far-fetched to some. But as we explore the cosmos, discovering interstellar objects such as ‘Oumuamua and the intriguing meteor IM1, we must consider the possibility that artifacts from other civilizations might have found their way to Earth. The vastness of the ocean, encompassing more than 70% of our planet’s surface, provides ample hiding spots for such relics to rest undisturbed for eons.
Recently, our research team embarked on an expedition to the Pacific Ocean, seeking to recover remnants of the first interstellar meteor, IM1. We have localized the meteor site to a kilometer squared region, increasing search efficiency by a factor of a hundred relative to past data. The material strength of IM1 was found to be orders of magnitude tougher than all 271 solar system meteors in the CNEOS fireball catalog. This begs the question: Could IM1 be a fragment of a broken Dyson sphere or another extraterrestrial artifact?
Finding extraterrestrial gadgets on the ocean floor would revolutionize our understanding of the cosmos. These relics could provide us with insights into the technologies, materials, and designs employed by advanced civilizations. Furthermore, the study of these cosmic artifacts could inspire humanity to strive for greater achievements in science, technology, and engineering, propelling us towards a future in which we take our place among the stars.
However, the search for extraterrestrial gadgets in the ocean depths is not without its challenges. The harsh conditions of the deep sea, including immense pressure, near-freezing temperatures, and complete darkness, require specialized equipment and expertise to navigate. Additionally, we must be prepared for the possibility that the artifacts we uncover might be beyond our current comprehension, necessitating a multidisciplinary approach to deciphering their secrets.
Despite these challenges, the potential rewards far outweigh the risks. As we plunge into the depths of our oceans, we also dive into the depths of knowledge. The discoveries we make will not only expand our understanding of the universe but also foster a sense of unity and purpose among humanity.
In conclusion, the search for extraterrestrial gadgets on the ocean floor represents an exciting new frontier in the quest for knowledge. The secrets that lie hidden beneath the waves hold the potential to unlock untold technological advancements and unite humanity in a shared pursuit of cosmic understanding. Just as Captain Nemo and his crew embarked on a voyage of discovery in the depths of the sea, so too must we venture into the uncharted waters of the cosmos, armed with curiosity and the determination to unravel the mysteries of our universe.”
GPT-4 did far better than GPT-3 on this task. I am flattered that GPT-4 is unanimously supportive of the Galileo Project’s plan to retrieve relics of IM1 and pays little attention to the pushback messages from some scientists on Twitter. A colleague alerted me today to some of the negative tweets about my latest published paper, titled: “Interstellar Objects from Broken Dyson Spheres”. In reply, I emailed back: “These tweets reflect more on those who wrote them than on anything else. Most theoretical astrophysics is about testing hypotheses. Consider the past decades in our search for the nature of dark matter. Why don’t the writers of these tweets ridicule the theorists who proposed Weakly Interacting Massive Particles (WIMPs) as dark matter? These WIMP speculations were already ruled out within the natural parameter space which was advocated for them. I do not see my discussions on the nature of interstellar objects as more speculative than WIMP proposals, and both were used to motivate experimental tests.”
Evidently, GPT-4 is more open-minded than some members of academia. This suggests changing the parent company name from OpenAI to Open-Minded AI.
Artificial intelligence holds the promise for being more open-minded than natural intelligence, and might therefore be better suited for advancing the frontiers of scientific knowledge while exploring the unknown.
The fundamental question is how much longer would scientists do better than machines? This brings a follow-up question based on the essay just written by GPT-4: will the superior machines in our future be of terrestrial or extraterrestrial origin?
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.
