Skip to content
Andrew Voirol
Thursday, November 12, 2026vitrail-build

Two-ply counterfactual Glazier AI, 520-seed passivity fuzzing, and chaos-tested Online Rooms

WebGPUArchitectureBenchmarkingShippingOpen Source

Because MuJoCo's mj_copyState clones full solver state into a scratch buffer with zero live world mutation, AI opponents no longer have to rely on static heuristics. I built Glazier AI (Pass & Play -> AI) using 2-ply counterfactual physics rollouts inside the Web Worker: first optimizing extraction pull azimuth via Cross-Entropy Method (optimalPullAzimuthRad), then scoring 2nd-ply adversarial top-course placements (twoPlyMasterScore01 = 0.763 on free center prisms vs 0.277 on pinned bearing columns) to maximize center-of-mass eccentricity for the next turn. Alongside the AI, I added a 520-seed × 80-action stateful trajectory fuzzer verifying T+U=mj_energyVel+mj_energyPosT + U = \text{mj\_energyVel} + \text{mj\_energyPos}T+U=mj_energyVel+mj_energyPos passivity with delta-debugging (ddmin) shrinking, plus C8 post-collapse sleep hardening (8/8 collapse seeds reaching 100% native tree sleep in 1.12–1.82 s with zero post-settle buzz). Finally, I shipped 5-character Online Rooms (?room=K7M4P) combining public WebRTC RTCDataChannel (< 4 KB binary 0x564d snapshots + round-robin 512-shard sync) with automatic <= 4.0 s ICE failover to a Cloudflare Durable Object SQLite WebSocket relay — verified across a 100-match adversarial network chaos simulator (8% drop + 65% burst loss, 120 ± 45 ms jitter, < 2.0 mm Guest convergence) and 58/58 Vitest invariants.

← Previous

Adversarial Colosseum audit: Maxwell-Fresnel optics, Timoshenko acoustics, and 3D shard rebound

Next →

Locking AI playfield art and faceted leaded stained glass within ±2 px


Andrew Voirol

Builder, hacker, shipper. Currently leaving localhost.

Navigate

WorkThreadsBuilder's LogAboutContactRSS Feed

Connect

X / TwitterGitHubLinkedIn

© 2026 Andrew Voirol·Back to top ↑
✦Just one prompt away from figuring it all out.