The Benefits of Systematic Over Anecdotal Evidence

Avi Loeb
6 min readMay 13


The “Burning Bush”, a 17th century painting by Sébastien Bourdon.

At the weekly Galileo Project meeting, the airline pilot and photographer Christiaan van Heijst, presented a photo of a mysterious cigar-shaped object which hovered over Spain in one of his flights a decade ago. This was an intriguing report of an anecdotal encounter with an Unidentified Aerial Phenomenon (UAP). Christiaan had visibility over a distance of about 10 kilometers. Airline pilots fly an average of 75 hours per month, namely about a tenth of the time. This means that such objects should appear once per year per 10 kilometers distance, ten times more frequently than Christiaan noticed because he monitors the sky only for a tenth of the time.

The Galileo Project (GP) that I have the privilege of leading, is engaged with a systematic search for UAPs. The first GP observatory is currently monitoring the sky above Harvard University out to a distance of about 10 kilometers. If the object noticed by Christiaan is a member of a population of similar objects that cover Earth uniformly, then by 2024 the GP observatory will record such an object in its data stream.

By making N copies of the first GP observatory and placing them in many geographical locations, GP could reach a statistical sensitivity per unit area and time that is N times better. This scaling is essential in removing optical illusions or instrumental artifacts which would be unique to particular instruments or particular locations or particular times. Moreover, there will be little risk from contamination by hackers if the locations of the observatories are unknown or move around.

There is a fundamental difference between anecdotal reports and a systematic search. Christiaan did not have state-of-the-art, well-calibrated scientific equipment at his disposal. He could not determine the distance to the object and therefore could not figure out its physical size or speed. Moreover, seeing one such object per decade does not allow for detailed follow-up studies.

However, a systematic search can pin down the statistics and physical properties of UAPs including precise distance measurements through triangulation from multiple GP sensors in the infrared, radio or audio bands. By using Machine Learning (ML) and Artificial Intelligence (AI), the GP research team plans to automate the detection procedure and not rely on human experiences. For the same reason, FIFA had used semi-automated offside technology at the Soccer World Cup 2022 in Qatar, offering a support tool for officials “to make faster, more accurate and more reproducible offside decisions.”

The 2022 UAP report from the Director of National Intelligence to the US Congress, suggested that about half of UAPs might be balloons. This could explain why the number of UAP reports correlates with population density. The more people monitor the sky, the more balloons they see. In selecting observing sites, GP should consider low-noise regions where the sky is less cluttered with balloons as well as regions with low population density but high rate of UAP reports.

A focus on anecdotal reports has the additional disadvantage that past events cannot be revisited in a short time so as to improve the quality of the data. For example, the DASCH project led by my colleague, Professor Josh Grindlay at the Harvard College Observatory, digitizes data recorded on photographic plates over the past century. The partial information on anomalous UAP from decades ago cannot be supplemented by higher quality data from modern telescopes and recording devices because unlike astronomical sources which stay in the same location in the sky, these UAP are gone by now. Similarly, the analysis of past reports from military personnel by the All-Domain Anomaly Resolution Office in the Pentagon, is limited in its scientific scope because of the anecdotal nature of past data.

This week, after all COVID-19 restrictions on campus were lifted, I held the first post-pandemic group meeting in my Harvard office. I asked the dozen members of my research group to briefly summarize what they are passionate about. The first two postdoctoral fellows of the GP research team, Richard Cloete from the University of Cambridge in the UK and Laura Domine from Stanford University, said independently that starting at a young age they dreamt of applying the scientific method to UAP. They were thrilled when I offered them a 3-year postdoctoral fellowship at Harvard University, since being part of such a project was their lifelong dream.

The transition from anecdotal eyewitness testimonies to systematic analysis of data from tens of GP observatories holds the promise of pinning down the nature of most UAP. The collection will likely be a mixed bag with most UAP having mundane explanations. But the fundamental scientific question is whether one or more of these objects is of extraterrestrial technological origin. If not, we could move on to explore this possibility farther away from Earth.

A systematic search offers the benefit of finding either functional devices or space trash — spanning a wide range of sizes. We already witnessed interstellar objects from the scale of a football field, like `Oumuamua, to the scale of a giant watermelon, like IM1. For a random distribution of interstellar objects, the meter-scale objects must be a million times more abundant than the hundred-meter-scale objects to explain their corresponding detection rates. The implied power-law in abundance versus size can be explained if small objects are pieces of bigger objects so that they carry a comparable amount of mass. The statistics are analogous to the abundance of pieces of different sizes from broken sea shells on the beach.

In addressing Enrico Fermi’s question: “Where is everybody?”, we should start from our home planet — namely UAP or IM1, move out to our back yard — namely anomalous interstellar objects like `Oumuamua within the orbit of the Earth around the Sun, and finally venture into interstellar space.

If we find any extraterrestrial technological product close to us, it would represent civilizations that had reached the third phase mentioned above. This implies that they are far older and potentially more advanced than we are.

The biblical awe of Moses while observing the burning bush not being consumed, would have been far greater if Moses had access to the GP assembly of multiple infrared cameras. The GP analysis of the bush data by the ML/AI software would have informed Moses about the effective temperature of the burning bush as well as the triangulated distance and power that it emits. Altogether, this scientific-quality data would have implied whether the burning bush is a UAP of natural or artificial origin.


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.



Avi Loeb

Avi Loeb is the Baird Professor of Science and Institute director at Harvard University and the bestselling author of “Extraterrestrial” and "Interstellar".