Video summary
DO NOT PLAY MW4 (Craziest evidence yet)
Main summary
Key takeaways
Overview of the Video’s Claims
- The video claims a new Call of Duty title is coming and argues players should understand what runs “behind the scenes,” not just what appears on-screen.
- It references an initial warning tweet alleging Call of Duty may access the user’s camera without a clear need (for example, while playing multiplayer).
Investigation Method (Technical)
- Two channels are credited for “digging into the files” that run alongside Call of Duty.
- The analysis focuses on monitoring which DLL (Dynamic Link Library) files a game process touches while running, using a tool that tracks runtime dependencies.
- DLLs are explained as shared code libraries that software loads to reuse existing functionality (which is normal in general).
- Therefore, the concern is not “does CoD use DLLs?” but which specific DLLs appear and why.
Key Findings Claimed
1) Baseline Check: Black Ops 1
- The video reports a small, “boring” set of DLLs appearing.
- It includes
ntdll, described as a standard Windows core compatibility dependency.
2) Modern Check: Modern Warfare 3 (More Suspicious)
- The DLL list is described as “exploding” (more system components showing up in newer games).
- The list is said to include
nvcamera64.dll, described as camera-related. - The video questions why a multiplayer shooter would need a camera DLL, implying interaction with camera/vision functionality.
Patent Connections Alleged
-
After finding the camera DLL, the video claims the investigation led to Activision patents involving facial imaging:
- A patent allegedly describing 2D facial images converted to 3D, including real-time tracking of a player’s face and expressions.
-
A second, more concerning patent is claimed to involve:
- Driving microtransactions in multiplayer games using engagement signals, including:
- player engagement metrics
- input speed/focus
- psychological states (e.g., frustration and satisfaction)
- and camera peripherals as one possible input source
- Driving microtransactions in multiplayer games using engagement signals, including:
-
The video argues these patents connect because:
- the facial imaging patent is cited inside the microtransaction patent, suggesting an intentional architecture that ties camera/face analysis to monetization.
Claimed Persistence + AI on Idle Resources
-
Another Activision document is cited describing:
- running/simulating after processing resources go offline
- distributing NPC simulations across network/user devices
- using spare capacity
-
The video’s interpretation:
- the game could use other players’ idle devices / leftover processes to run AI.
- it frames this as AI trained on real human behavior (matches) to make bots more convincing.
-
The video also claims some players found processes that stayed active after the game was closed, connecting this to lingering simulation/processing.
“Putting it All Together” (Overall Thesis)
-
The video concludes that individually plausible items become concerning when combined:
- camera-related DLL presence
- facial imaging patent
- engagement/frustration-based microtransaction patent (with camera peripherals cited)
- idle-device simulation/AI persistence concept
-
It argues this could enable games to learn more about the person behind the screen (such as frustration, engagement, and spending behavior) and potentially adapt matchmaking/content/offers in real time.
-
The creator also claims this pattern is not observed in another shooter:
- XDefiant is described as using a similar live-service engagement/monetization model, but without the camera DLL appearing in its process list.
Reviews / Guides / Tutorials
- This is framed as a “how the investigation works” walkthrough covering:
- how to monitor DLL usage at runtime
- using DLL sightings as leads to search patents
- correlating systems (camera → facial analysis → engagement/microtransactions → AI/simulation concepts)
Main Speakers / Sources
- Main speaker: the unnamed video narrator/host (delivering the investigation and conclusions).
- Digging sources: credited as two “channels” described as “legends” (specific channel names are not provided in the subtitles).