Hands-on Gemini 4 Neon: SDF dragon, walkable house from a 1959 floor plan, no-training neural nets
Prompt Engineering · youtube · 2026-11-19
The Prompt Engineering YouTube channel got early access to Gemini 4 Neon and skips benchmarks in favor of real outputs:
- SDF dragon as one formula, echoing Keenan Crane's experiment
- One-file animated SVG peacock, and a walkable 3D house generated from a 1959 floor plan
- Agent swarm building a voxel diorama; an on-call agent deployed on Cloud Run, then broken and rolled back
- Three DGX Spark GPU experiments: Growing Neural Cellular Automata (Distill 2020), a four-legged robot gait, colliding galaxies
- A neural network written with no training, via MIT CSAIL weight-space programming
Also covers Antigravity's four thinking levels and three rules for long tasks. The author recently joined Google; views are his own.
More from coding & agent
- AI is letting hackers attack at scale, warns security pro amid week of outages — IanAndrewsDC · 2026-11-19
- Dev finds hacky way to gate agent flows with verifiable human authorization — csuwildcat · 2026-11-19
- Code is a liability, not an asset: AI coding makes restraint more valuable — bendee983 · 2026-11-19
- Gemini 4 Neon's 39% SWE-sweep questioned: results hinge on the harness — OfirPress · 2026-11-19
- Why the most AI-pilled engineers cluster at labs: unlimited tokens, agents, and humans in the loop — dbreunig · 2026-11-19
- Harness is a hidden variable: researcher blasts AI labs as terrible engineers — gerardsans · 2026-11-19