Powerful AI uncovers over 100 hidden exoplanets in NASA’s dusty archives, raising questions about government-funded science wasting taxpayer dollars on manual drudgery while ignoring earthly priorities.
Story Highlights
- University of Warwick’s RAVEN AI validated 118 exoplanets, including 31 new ones, from TESS data, plus over 2,000 candidates.
- Discovery reveals rare ultra-short-period planets orbiting in under 24 hours and quantifies the “Neptunian desert” at just 0.08% occurrence.
- TESS matches Kepler’s capabilities with 10x better precision, accelerating discoveries from years to months.
- Researchers emphasize AI’s objectivity in sifting massive datasets, freeing resources for deeper studies like JWST follow-ups.
RAVEN AI Validates Hidden Worlds
Astronomers at the University of Warwick deployed RAVEN, an AI tool trained on simulated data, to analyze the first four years of NASA’s TESS observations. TESS monitored 2.2 million stars since 2018, detecting exoplanets through transit photometry—dips in starlight from planetary passages. RAVEN distinguished true transits from false positives like eclipsing binaries. The effort validated 118 exoplanets, with 31 newly confirmed, and identified over 2,000 high-quality candidates, nearly 1,000 previously undetected. This focused on close-in planets orbiting in under 16 days.
Rare Planetary Types Emerge
RAVEN spotlighted extreme worlds, including ultra-short-period planets completing orbits in less than 24 hours. It also probed the “Neptunian desert,” a gap in Neptune-sized planets near stars caused by photoevaporation. Kaiming Cui’s analysis pegged this rarity at 0.08% around Sun-like stars—the first precise measure. RAVEN confirmed 9-10% of such stars host close-in planets, aligning with Kepler baselines but with uncertainties up to 10 times smaller. Dr. Marina Lafarga Magro called this one of the best-characterized samples of close-in planets.
Expert Insights on AI Efficiency
Dr. David Armstrong, senior team member, stated RAVEN analyzes enormous datasets consistently, enabling reliable population mapping. Dr. Andreas Hadjigeorghiou highlighted AI’s superior pattern recognition over human review for mimics. The March 2026 publication in Monthly Notices of the Royal Astronomical Society underscores peer-reviewed rigor. No major contradictions exist across reports, though headlines rounded “118” to “100+.” Candidates await mass confirmation via radial velocity, paving way for JWST atmospheric studies.
Implications for NASA and Taxpayers
Short-term, these finds boost NASA’s exoplanet catalog—currently around 6,000—by about 2% if candidates confirm. Long-term, RAVEN enables habitability models and archive re-mining for hundreds more worlds. This shifts astronomy from slow manual validation to automated pipelines, applicable to future missions like Rubin Observatory. Efficient NASA spending aligns with fiscal conservative priorities, maximizing taxpayer value amid frustrations over government overspending. Yet, as both sides lament elite distractions, earthly crises like inflation demand equal scrutiny.
Sources:
100 new alien worlds: Scientists find hidden haul in data from NASA exoplanet-hunting spacecraft
Artificial intelligence found dozens of hidden planets in NASA data
AI Approach Uncovers Dozens Of Hidden Planets In NASA’s TESS Data
Over 100 hidden planets discovered in NASA data: artificial intelligence at the service of astronomy
