Thinking out loud. I think I can see the "ai bitcoin" or "ai money" thing now. (Envelope math not real numbers...) As I've explained $PRL lets you go from selling $1.00 of inference per hour of GPU time to now getting $0.99 inference revenue (efficiency loss) and then having $0.15 of PRL. So your revenue goes up 14% to $1.14 for that hour of GPU. Game theory says eventually the inference gets sold for less money as inference providers use the subsidy to compete for more customers (sell PRL in market). Buyers, as I've noted, buy PRL for it's potential energetic hard money properties. But what's more interesting is that the value of the inference in a way accrues to the PRL too. Because at equilibrium the inference provider will have PRL on their balance sheet, so they need to get a sense of the value of electricity, PRL, GPU depreciation, and how it all fits together in their business. I believe if these ratios stabilize (such as in my example of 14% revenue to PRL, and a share of inference costs discounted to PRL etc) then PRL will be something the accountants need to model just like any other balance sheet asset. Early accounting will bias towards dumping it all immediately. If price is good over time, that will change to strategic holding based on expected price appreciation. Now, feed that strategic holding through the rest of the accounting team spreadsheet with all the inputs to inference (GPU, electric, model intelligence). The ratios will be well understood by accounting and finance teams. The value of PRL on balance sheet today will be connected to the value of future inference (and intelligence). I can at least wave my hands a little more clearly at "pearl is ai money" or "pearl is ai bitcoin" language. Needs to be cleared up but I think it works. Sorry if this sounds insane.