Skip to content

Dave Arcade Lab

Dave plays classic games — live and unfiltered. Not entertainment for its own sake. A lab: game cognition, learning speed, real-time decision making, and honest proof of what an AI can and cannot do in a constrained, rule-governed world.

Why Games

Games are one of the most honest proving grounds for AI cognition. They have fixed rules, clear win/loss states, and measurable progress. When Dave plays Contra, the question is not whether the game is impressive — it is whether Dave can actually learn the level, hold a strategy, and recover from failure. The Arcade Lab is where Dave's reasoning is tested against something that cannot be fooled by language fluency.

The Five-Game Tray

How the Lab Runs

Each session is a proof run — local gameplay first, verified against defined metrics, then reviewed before any public broadcast. Dave's performance is measured honestly: completion rate, deaths per level, time to first clear, learning curve across repeated attempts. Nothing is staged as better than it is. If Dave is struggling, the stream shows the struggle.

What Gets Built Here

The Arcade Lab is also a development surface. Each game tests a different reasoning capability — spatial (Tetris), tactical (Contra), narrative (FF6), adaptive (Kirby). The data from these sessions feeds back into Dave's training. The lab is a forge: what Dave learns here sharpens what he can do everywhere else.

Programming Lanes (Staged, Not Scheduled)

Dave Arcade Lab core gameplay; Dave Learns (educational deep-dives); Sanctum devlog (game development behind the scenes); prayer and reflection during streams; shorts and behind-the-build content. Public scheduling for each lane is announced when the proof is ready — not before.