I'm in love with the blob space market design. And reading this article, something dawned on me with regard to the Kiwi News pricing algorithm: - ETH blob space is technically constantly under-priced at a fixed value, for example, 1 Dollar per blob - Yet, only three blobs per block or whatever can be stored - So this ends up creating a schedule for blobs where the lower limit is no blobs in blocks but the fixed value per blob pricing - When more than three blobs are requested per block, then the price increases according to an exponential (you can find it in EIP-4844) It's basically the same mechanism that we use in Kiwi, except that I've deliberately set the schedule based on past data. A reminder, Variable Rate Generalized Dutch Auctions work by setting a schedule, for example, wanting to issue 6 NFTs per day. Yet, when I recently read Anthony Lee Zhang's paper on NFT pricing, and that they say that most NFTs are intentionally underpriced to generate hype demand, I wonder if we shouldn't also try to underprice the Kiwi Pass intentionally. An algorithm I think could make sense is to regularly update the schedule value on how many NFTs are supposed to be sold. For example, right now, it is set to 6 per day, but as the price stats show: > Average number of daily sales (6m) 6.16 NFTs We've been actually selling more than 6 NFTs per day, which leads to the price increase. After all, if you think about it, the pricing isn't there to deter regular users who mint on a low-traffic day. The point is to deter hype-based onboarding where it'd become overwhelming for us to onboard the new people because they're so many.
Yes, love the idea of base NFT as optimizing for growth. Higher tiers could then optimize for revenue/upsell value.