Citizen Scientist Uses AI Tools Like Claude Code to Spot Potential New Planet in NASA Data

The rapid evolution of artificial intelligence over the past few years has shifted the technology from a futuristic novelty into an accessible toolkit for everyday individuals. While early iterations of large language models were largely celebrated for generating creative text, humorous images, or unconventional videos, today’s advanced AI systems are empowering everyday people to tackle complex scientific and technical challenges that were once reserved exclusively for seasoned experts and institutional researchers.

Earlier this year, the tech community watched as open-source enthusiasts leveraged AI models to build comprehensive, free software suites as alternatives to expensive commercial software like Adobe. Now, the boundaries of citizen science are being pushed even further. A space enthusiast has reportedly used advanced artificial intelligence tools, including Claude Code, to help identify a potential new exoplanet hidden within publicly available NASA data. While the discovery remains unconfirmed and awaits formal validation, the milestone highlights the growing synergy between modern AI models and independent scientific inquiry.

Claude Code May Have Just Discovered a New Planet for NASA

Huge Emphasis on "May"

The remarkable journey began when Pavel Rabtsevich, an enthusiast with a deep passion for astronomy and modern technology, decided to comb through public astronomical datasets provided by NASA. While analyzing light curves and stellar activity, Rabtsevich noticed an intriguing anomaly: a specific star appeared to dim rhythmically, dropping in brightness for roughly a two-hour window every three days.

Suspecting that the recurring dimming pattern might be caused by an exoplanet transiting—or passing directly in front of—its host star, Rabtsevich decided to dig deeper. He cross-referenced his initial findings with additional archival data to ensure the phenomenon was consistent and not merely a glitch or background noise in the instruments. The repeating dip in starlight remained robust, prompting him to seek advanced computational assistance to analyze the complex telemetry.

As Rabtsevich later detailed in posts shared on social media, he began feeding the astronomical data into several sophisticated artificial intelligence platforms, including Claude Code, OpenAI Codex, and Jev, to evaluate whether his hypothesis held statistical weight. By prompting the AI models to rigorously analyze the data and iterate through various parameters, he eventually deduced that the target star was cataloged as TIC 4206066. According to the AI-assisted analysis, the data suggested the presence of a potential exoplanet roughly 1.4 times the size of Earth, completing an extraordinarily tight orbit around its host star every three days, which translates to a remarkably short planetary year.

Determined to test his theories thoroughly, Rabtsevich utilized Claude Code to run a series of 74 distinct computational and statistical tests designed to challenge the exoplanet hypothesis. The AI system processed the extensive datasets, stress-tested the assumptions, and ultimately concluded that Rabtsevich’s planetary transit theory was mathematically sound. To further validate the model’s predictive power, Rabtsevich concealed specific segments of data from the AI and tasked the system with forecasting stellar behavior during those missing gaps. Claude Code successfully anticipated when the star’s light would dim, strengthening the credibility of the underlying calculations.

The next critical phase of the process involves official scientific confirmation through NASA and astronomical institutions. To advance the discovery toward formal verification, Rabtsevich submitted a request to the Transiting Exoplanet Survey Satellite (TESS) project, asking researchers to direct the space telescope’s instruments toward the precise coordinates of TIC 4206066. The request was accepted and is currently integrated into the queue of scheduled TESS observation tasks.

As the space telescope passes over that designated region of the cosmos, astronomers and researchers will gather fresh photometric data to determine whether the star exhibits the predicted transit dips. Because Rabtsevich has already mapped out the exact dates and times when the stellar dimming should occur based on his models, verifying the discovery will hinge on whether upcoming TESS observations match the computed timeline. If the space telescope’s data aligns with the predictions, Rabtsevich will move one step closer to making a verified astronomical discovery.

The intersection of advanced AI coding assistants, public space agency data, and independent citizen scientists points toward a shifting landscape in modern research. What began as an experiment with generative media tools has matured into a practical avenue for everyday enthusiasts to contribute meaningfully to scientific exploration. As artificial intelligence continues to advance, the boundaries of who can participate in major scientific breakthroughs are expanding faster than ever before.