I love Airtableโme and Mindvalley founder Vishen Lakhiani have been recommending Airtable to tens of thousands of our customers for years... and I included a lesson about Airtable AI into our "AI Mastery" program, however, the way your AI credits system works makes it difficult to recommend Airtable AI to anyone.
Most probably your system uses credits every time you generate previews on "Long text" fields which is a very flawed UX design and/or unethical billing practice... because:
โข (A) You set GPT-4o as the default model (which is an overkill in 90%+ cases) and you don't allow users to set "GPT-4o mini" as the default. GPT-4o burns through credits like there's no tomorrow.
โข (B) You automatically generate previews (without asking for a user approval) and burn through their credits.
โข (C) You do not state in any documentation and you do not provide any warnings that AI credits are being used during test/preview runs.
As a result of the above I received dozens of complaints from my students who burnt through their AI credits after using Airtable AI to fill in only a few cells during our training! This is nuts.
Fix this ASAP. Billing issues are no joke.
p.s. I'll soon deliver Airtable AI training to Mindvalley employees. As a result of that training we may purchase hundreds of additional AI Add-ons... but if you don't fix the above I'm afraid I'll not be able to approve purchasing the additional licenses.
