Home of DISCERN-AI, the deepfake detection system that asks the one question a forgery can’t answer: Is anything living behind this?
Found Awardable twice in 2026. For the Resonant Sight System (Project Uriel) and for MTCR-119, the Möbius Toroidal Coherence Resonator.
Read the announcement →Listen or watch, make your call, then see what DISCERN-AI said. Every clip below comes from a public benchmark, and every score is the one DISCERN-AI actually gave it.
How these clips were chosen: 24 voice and 24 video clips were drawn at random from the full benchmarks, using a seed fixed before the draw. None was swapped out afterwards. On this pool DISCERN-AI calls 24 of 24 voice clips and 22 of 24 video clips correctly. Video scores were produced while each clip was held out of training. The vertical line on the score bar is the benchmark’s decision threshold. Two explanatory videos are in the video library.
DISCERN-AI benchmark edition of 9 November 2026. Every number is from a complete run. Full results document (PDF)
13 AI voice engines, none of them seen in training. Eleven are caught at 95.9% or better. The two hardest are shown openly: A17 at 25.0% and A18 at 85.6%.
Each generator is held out of training in turn and then tested. AUC per held-out test (1.0 is perfect separation, 0.5 is chance).
Early in 2024, a finance employee at the engineering firm Arup wired $25.6 million during a video call in which every colleague on screen was synthetic. DISCERN-AI was built to stop that call.
Most detectors chase the fingerprints a particular generator leaves behind, so every new release can break them. DISCERN-AI measures something else: the closed-loop signature of a living body, where breath, voice and motion keep adjusting to one another. On generators it has never seen, it still holds: Sora at 0.986.
One exclusive partner per vertical. All rights outside that vertical are retained.
Protected by U.S. Non-Provisional Patent Application 19/700,318, pending (priority 13 June 2025).
Passive space-domain awareness. Priority 13 June 2025, pending.
Anti-periodic boundary conditions in a driven plasma. Concept stage, pre-prototype.
A drive architecture with no gearbox, no mechanical differential and no routine friction braking.
The second patent application, anchored at 41.176 Hz. Open the live Klein Spiral instrument →
The research program behind all of it, including the cancer-as-coherence-collapse white paper and the Klein Spiral.
The announcement of both Awardable findings.
Paul Samuel Guarino is an independent researcher and inventor working across signal detection, environmental monitoring, and applied electromagnetics. His work is organized around a single derived constant — 700 ⁄ 17 ≈ 41.176 Hz — applied as a reference anchor across otherwise unrelated detection problems.
He is the director of the research, not its credentialed authority: he sets the claim, chooses what gets tested, and defines the standard a result has to clear. The engineering and the verification are done in the open, and a negative result is reported as plainly as a positive one.
Why DISCERN-AI exists, how it works, and the benchmark proof: video, voice and face swaps.
Inside the patent-pending Vimana CP-17 drive architecture.
On AI risk, self-regulation, and who bears the consequences.
Archaeoacoustic resonance and the 700/N cascade.
On AI, the people who make it, and what creation obliges.
Benchmark clips used in the demo above, trimmed to 4 seconds and re-encoded for scoring: AIGVDBench — L. Ma et al., “Your One-Stop Solution for AI-Generated Video Detection,” CVPR 2026, CC BY 4.0, whose real videos come from OpenVid-1M (CC BY 4.0). ASVspoof 2019: M. Todisco, X. Wang, V. Vestman, M. Sahidullah, H. Delgado, A. Nautsch, J. Yamagishi, N. Evans, T. Kinnunen, K. A. Lee, “ASVspoof 2019: Future Horizons in Spoofed and Fake Audio Detection,” Proc. Interspeech 2019, ODC-BY.