This post was originally published on here
In the quiet suburbs of Pasadena, California, an 18-year-old high school senior named Matteo Paz turned a summer research stint into a astronomical breakthrough that has astronomers buzzing. By developing a sophisticated machine-learning algorithm, Paz sifted through vast archives of data from NASA’s retired NEOWISE telescope, uncovering 1.5 million potential new celestial objects. This feat not only earned him a $250,000 prize in the Regeneron Science Talent Search but also highlighted the growing power of artificial intelligence in unlocking secrets of the universe.
Paz’s journey began with a childhood fascination for the stars, sparked by family outings to public astronomy lectures. As a high school student, he secured an internship at the California Institute of Technology (Caltech), where he worked under the mentorship of astronomers like Dmitry Duev, a former University of Arizona PhD who now contributes to Caltech’s efforts. There, Paz delved into the underutilized datasets from NEOWISE, a mission originally designed to hunt for near-Earth objects but which amassed billions of infrared observations over its decade-long operation.
The challenge was immense: NEOWISE’s archive contained over 200 billion data points, many of which had been overlooked due to the sheer volume and the limitations of traditional analysis methods. Paz recognized that manual review was impractical, so he set out to build an AI tool capable of processing this deluge at unprecedented speeds. His algorithm, trained on machine learning principles, could analyze images in sub-millisecond times, identifying anomalies that might indicate previously unknown asteroids, stars, or other cosmic phenomena.
The Algorithm’s Inner Workings
At the heart of Paz’s innovation was a neural network designed to detect faint signals amid noisy infrared data. Drawing from techniques in computer vision, the model learned to distinguish between static stars and moving objects by comparing sequential images from the telescope’s scans. This approach allowed it to flag potential discoveries that human astronomers might miss, effectively automating a process that once required painstaking manual verification.
Paz’s work didn’t stop at detection; he also ensured the algorithm’s outputs were verifiable. By incorporating validation steps, such as cross-referencing with existing catalogs, he reduced false positives and provided a framework for other researchers to build upon. The results were staggering: in just six weeks, the AI processed the entire dataset, revealing 1.5 million candidates for new astronomical objects, many of which could be asteroids or distant galaxies.
This discovery has implications far beyond a high school project. As detailed in a peer-reviewed paper Paz authored solo in The Astronomical Journal, the findings expand our understanding of the solar system’s inventory and could aid in planetary defense efforts by identifying potential Earth-threatening asteroids. The paper, published earlier this year, underscores how accessible tools like Python programming and public NASA datasets can democratize scientific research.
Mentorship and Institutional Support
Paz’s success story is also one of effective mentorship. At Caltech, he collaborated with experts who provided guidance on both the astronomical and computational aspects of his project. Dmitry Duev, in particular, helped refine the algorithm’s parameters, drawing from his experience in machine learning applications for astronomy. This collaboration exemplifies how academic institutions are bridging the gap between professional researchers and young talents.
The Regeneron Science Talent Search, often dubbed the “junior Nobel Prize,” recognized Paz’s contributions with its top award in March 2025. According to reports from Smithsonian Magazine, the competition judges were impressed by the real-world applicability of his AI model, which processes data at speeds unattainable by previous methods. This win not only provided financial support for Paz’s future education but also spotlighted the role of youth in advancing fields like astrophysics.
Beyond the prize, Paz’s project has sparked discussions in the scientific community about integrating AI more deeply into observational astronomy. Forums on platforms like X (formerly Twitter) have buzzed with posts praising the teenager’s ingenuity, with users noting how his work echoes historical discoveries like Clyde Tombaugh’s finding of Pluto in 1930, but accelerated by modern technology. One popular thread highlighted the potential for similar AI tools to analyze data from ongoing missions like the James Webb Space Telescope.
Broader Implications for Astronomy
The integration of AI in astronomy is not new, but Paz’s achievement demonstrates its scalability for large datasets. Traditional methods often bog down in the petabytes of information generated by telescopes, leading to backlogs of unanalyzed data. Paz’s algorithm addresses this bottleneck, offering a blueprint for future projects that could uncover even more about the universe’s composition.
Experts point out that many of the 1.5 million objects could be main-belt asteroids or distant stellar remnants, contributing to catalogs that help model the evolution of our galaxy. In a feature by Phys.org, astronomers at Caltech emphasized how Paz’s discoveries broaden the scientific yield of the NEOWISE mission, which retired in 2020 after providing invaluable infrared surveys.
Moreover, this work aligns with NASA’s push for open data initiatives, encouraging citizen scientists and students to contribute. Paz’s story inspires a new generation, showing that with computational skills and curiosity, significant contributions are possible without access to cutting-edge observatories. Recent news on X reflects this sentiment, with educators sharing how advanced math and programming empower students to tackle real-world problems, much like Paz did.
Challenges and Future Directions
Despite the triumphs, Paz’s project faced hurdles. Training the AI required handling noisy data, where cosmic rays and instrument artifacts could mimic genuine signals. He iterated through multiple model versions to improve accuracy, a process that involved rigorous testing against known objects. This iterative approach is a hallmark of robust AI development, ensuring reliability in scientific applications.
Looking ahead, Paz plans to pursue studies in astrophysics or computer science at a top university, potentially building on his algorithm for other datasets. The scientific community is already adapting his methods; for instance, researchers are exploring its use in analyzing data from the European Space Agency’s Gaia mission, which maps billions of stars.
The broader field is witnessing an AI revolution, as noted in a December 2025 article from Cyprus Mail, which discusses how machine learning is accelerating the search for exoplanets and extraterrestrial intelligence. Paz’s contribution fits into this narrative, proving that innovative algorithms can reveal hidden patterns in cosmic data.
Industry Perspectives and Ethical Considerations
For industry insiders in tech and aerospace, Paz’s work underscores the value of cross-disciplinary skills. Companies like SpaceX and Blue Origin are increasingly incorporating AI for mission planning and data analysis, and talents like Paz represent the future workforce. His algorithm’s efficiency—processing billions of entries in weeks—highlights potential cost savings in research budgets.
Ethically, the rise of AI in astronomy raises questions about data ownership and verification. While Paz’s open-source approach promotes collaboration, ensuring that AI-driven discoveries are peer-reviewed remains crucial to maintain scientific integrity. Discussions on X have touched on this, with posts warning against over-reliance on algorithms without human oversight.
Paz himself advocates for accessible education in AI and astronomy, sharing in interviews how public resources fueled his project. As reported by SciTechDaily, he hopes his success encourages more students to explore these fields, potentially leading to more breakthroughs.
Personal Reflections and Legacy
Reflecting on his journey, Paz credits his family’s support and early exposure to science as pivotal. Growing up, stargazing lectures ignited a passion that evolved into a tangible impact on astronomy. His single-author paper stands as a testament to individual ingenuity in an era of team-based research.
The legacy of Paz’s discovery extends to planetary science, where identifying new objects could refine models of solar system formation. Astronomers anticipate that follow-up observations with ground-based telescopes will confirm many of his candidates, adding to our cosmic inventory.
In the grand scheme, stories like Paz’s remind us that innovation often comes from unexpected quarters. As AI continues to permeate scientific inquiry, the boundaries of what’s possible expand, promising a future where teenagers might routinely contribute to mapping the stars. His work, celebrated across media from Smithsonian Magazine to viral X posts, signals a shift toward more inclusive and accelerated discovery in the cosmos.
