Very interesting 36 hours. Monday morning I get an email from Yelp, saying my free trial of the Yelp API was over in 24 hours and they wanted $220/month if I wanted to keep using it. I like my little AI toy but I don't make it that much. So I flew to Vegas for work for a meeting on Monday and Monday night in the terminal, with nothing to do but wait for my flight, I did the following. 1. Ask the AI god which other APIs would be good, and we concluded Google Places API would be best. 2. The API is not free, but there's no charge if you have under $200/month of usage 3. I told the AI god to build a caching capability, so that I would only pull in the dive bars for a city at most once per week, and just cache(database) if the last request was less than a week old. This ensures that I will never hit the $200 limit. 4. The Google API did NOT have the meta parameter "divebar" like Yelp had, so I when I pull all the bars into the cache the first time, the AI god reads the reviews from customers and decides if it is a dive bar or not and sets a dive part attribute as Y or N. I did all of these changes in one hour and half of that was dicking around with Google Console, figuring out how to turn on the right API. The site is now better as the AI generated Dive Bar designation is better than the meta data version from yelp. I am not kidding, i did this in one hour. At work, this would have taken two hours to WRITE THE JIRA ticket and then engineering would have spent 2 or 3 weeks on this. I am so bought into this AI thing. This is going to condense three jobs into one(Designer+engineer+product manager). Also, it did my SEO work for me in five minutes and I am fifth on Google now for "dive bar finder" ** Edited by MrBayAreaHokie at 2/17/2026, 9:52:45 PM