I had the same feeling. I think it would click more if the cart visibly changed after each station — add a layer, change color, get cut into dies, etc.
Then the animation would be doing some of the explaining instead of mostly being scenery.
1 karma · joined August 7, 2026
Then the animation would be doing some of the explaining instead of mostly being scenery.
If the weights are physically encoded in hardware and the attacker owns the device, the problem becomes hardware extraction: decapping, probing, imaging, side channels, etc.
You can make that very expensive, but it’s still a very different security model from keeping the weights in a datacenter.