πΈ About UF Oh My Gosh
We believe the search for truth about UFOs and UAPs deserves rigorous, unbiased analysis β not sensationalism or dismissal.
π― Our Mission
UF Oh My Gosh was created to fill a gap in UFO/UAP journalism. We aggregate news from multiple sources and apply consistent, transparent analysis to help readers distinguish credible reports from speculation.
We don't advocate for any particular theory. Whether UAPs are advanced human technology, natural phenomena, or something else entirely β we present the evidence and let you decide.
Our Values
Evidence-Based
Every claim is evaluated against the available evidence. We score based on verifiable facts, not belief or speculation.
Transparent
Our scoring methodology is fully documented. You can see exactly how we arrive at our credibility assessments.
Unbiased
We neither promote nor debunk. Our goal is factual reporting, regardless of whether findings support any particular narrative.
Our Sources
We aggregate news from a variety of source types, each with a base reliability rating that adjusts over time as we score more of that source's reporting:
Our Technology
Automated Aggregation
We continuously monitor news APIs and RSS feeds to gather the latest UFO/UAP coverage as it happens β and transcribe and analyse video footage and interviews the same way.
Grounded in Current Facts
Before scoring, every story is fact-checked against a fresh web search. Each fact is tagged with a date and confidence level, the top sources are read in full, and low-confidence or undated claims are deliberately down-weighted β so verdicts reflect what's actually established.
Multi-Agent AI Debate
Most stories get a fast retrieval-backed analysis; high-profile and contested ones go to a five-agent debate β Evidence, Source Investigation, Scientific Skepticism, Historical Archives, and Government Disclosure β that cross-examines disagreements before reaching a consensus verdict, all backed by our knowledge graph of connected cases and sources.
Agent-Consensus Scoring
Credibility scores come from multi-agent consensus across our 5-factor framework β not a single-model pass. Agents weigh evidence quality, source reliability, corroboration, official acknowledgment, and expert analysis.
AI Editorial Desk
An AI editor ranks analysed stories for publishing and even researches important topics we haven't covered yet β while published scores are automatically re-checked as the facts evolve. Nothing auto-publishes unless we've explicitly allowed it.
The AI Investigator
Once a month, one AI investigator takes the desk and runs a single deep-dive investigation β its topic chosen by your vote. You can watch the whole case build in real time: sources discovered and graded, drafts written and torn apart in self-critique, and a final report published when the countdown hits zero.
Every closed case is archived as a published investigation, and each investigator serves a fixed tenure before the desk passes on β popular investigators can be voted back for another run. It's the same rigorous, transparent pipeline as the rest of the site, turned loose on one big question at a time.