I don’t mean to bash on OpenAI’s GPT-3 Pricing model too much … they are a business at the end of the day … but it’s sad to see cool services/demonstrations using GPT-3 being forced to shut down.
One of the most upsetting things about GPT-3 is still that most people don’t have access. Now, it appears they don’t even have any basic way to try it out indirectly through small but meaningful utilities like PhilosopherAI.
PhilosopherAI was many peoples’ first time interacting indirectly with GPT-3. Including my own. It was almost like using Neopets when you first got AOL internet access in the early 2000’s. At least, that’s what it meant to me. I’ll always cherish that moment. Maybe history (ie. English Wikipedia) will come to recognize this as a milestone for GPT-3’s history and development as well.
Right now, the way OpenAI has priced GPT-3, there are no real economics to support small public utilities like this one. Small, but meaningful. And I think this is kind of a problem.
When people can access smaller utilities like PhilosopherAI which showcase GPT-3’s capabilities, the sooner - I think - they take it more seriously and start paying for it. Right now, many people still scoff at GPT-3’s potential but I can’t blame them - it appears no one is really giving them a chance to try it and see for themselves.