Others have speculated that the API pricing model is built around customers who want to use the data for AI training, not customers who want to build apps for public use. The $20M price tag is what they’re hoping a mega corp will pay for data access and don’t care about anyone who can’t afford that much. Some money is better than no money, but for a lot of people the “chance” at BIG money is better than some money lol
If this is the case, I don’t understand why they wouldn’t just separate into tiers, where mass data usage to feed into a language model is priced differently than people legitimately using and contributing content to the site.