The problem of this "depreciation" rationale is that it presumes that all the cost is in training, ignoring that actually serving the models is also very expensive. I certainly don't believe they would be profitable, and vague gestures at some hypothetical depreciation sounds like accounting shenanigans.

Also, the whole LLM industry is mostly trying to generate hype, at a possible future where it is vastly more capable than it currently is. It's unclear if they would still be generating as much revenue without this promise.