This Camera Mints an NFT with Every Shot
Deep fakes and AI-generated images are now commonplace. Spend more than a few minutes on social media and you’re bound to come across completely fake photos of politicians and celebrities. And though most people are aware of that possibility, they can’t always spot the fakes — especially when those fakes happen to align with their own preconceptions. We desperately need a solution to this problem, because it will only get worse as the technology improves. Human Controller developed one potential solution in the form of a special camera that mints an NFT with every shot.
You probably know NFTs (non-fungible tokens) as those silly crypto ape images that people tried to pass off as investments a few years ago. But while that use case was ridiculous, NFTs can serve a legitimate purpose. In theory, an NFT is a unique virtual item authenticated by an entry in a blockchain. When handled properly, an NFT can be proven to be the original. This camera takes advantage of that to provide assurance. Any photo taken with the camera matchesa traceable NFT. If someone wants to know whether that photo represents reality (at least as the camera saw it), theycan verify the authenticity like they would with any other NFT.
The NFT Camera looks absurd, which is part of the point. Nobody is going to mistake this for another camera. The “body” of the camera is made up of giant letters spelling “NFT.” The only visible feature on the front is the small lens hole. The back side has an LCD screen that shows the current frame and a big “Mint now” button. Pressing that button snaps a photo. The user then sees their photo and pressing the button a second time mints the NFT in the Astar Network blockchain, tied to the selected crypto wallet. The image itself isn’t stored in the blockchain — just a unique identifier.
The hardware includes a Raspberry Pi 4 Model B (2GB) paired with a Raspberry Pi Camera V2. The display is a 3.5″ touchscreen and power comes from a USB battery bank. All of the functionality is software.
But there are a few problems with this concept. The most obvious is that a user could simply print out a fake photo and then use this camera to take a picture of that. A nefarious user could also spoof the whole process to mint NFTs tied to fake images. Human Controller suggests a solution to the first issue, which would work like a smartphone’s facial recognition unlocking feature to ensure it is looking at a 3D world and not a 2D image. The second issue would be much harder to overcome and would require some way to lock down cameras, with people only trusting NFTs minted by such cameras.