I ve been testing a safety check for an AI app by running a script of test messages many times against the model. Every run was fine until it wasn t: after a day of repeated runs, calls started failing with 429 errors, and waiting 60 seconds between retries stopped helping.

I assumed the limit was per minute. It wasn t. On the free plan for the model I use, the limits are 30 requests a minute, 1,000 requests a day, 8,000 tokens a minute and 200,000 tokens a day. I believe I hit the daily cap, because the app itself also stopped generating responses at the same time. My tests and the app share one account, so they share one budget.

What I m changing: run tests in small batches, never repeat full runs, and plan a backup provider so a quota problem doesn t become an outage.

How do you keep testing from eating your production quota? Separate accounts, a cheaper model for tests, or something else?

"I saw @busmark_w_nika mention in another thread that LinkedIn now scans AI-generated posts and may reduce reach for content that 'sounds the same.'

This got me curious I use AI for drafting but always rewrite the hook myself. Have any of you actually seen a drop in reach when using AI drafts vs. writing from scratch?

My longest experience started in a startup with @minimalist phone: reduce your screentime . (originally a digital detox app for Android, then iOS).

From a marketing perspective, the following worked best for us (almost always):

• paid ads

• UGC videos (also as paid ads)

There s a small difference in how you ask an AI to save something, and it changes what gets saved.
Save this for next time. The AI writes a message to the next AI, and it comes out as instructions:
Next session: confirm the hotel before booking trains. Do not change the Paris dates. Remind the user about the deposit.
Save what we settled. The AI writes down facts:
Paris is three nights. Trains between cities; a rental car was dropped because of city parking. The hotel deposit is due May 1.
The second note stays true after the conversation ends. A month later, it still says where things stand and why.
The first is a set of instructions from a moment that has already passed. A new conversation reads them and follows them, even after the plans have changed.
I saw this in my own records.
Notes written as instructions carried an earlier stage into new conversations. Notes written as facts read like a record: what was decided, and why.
Now when I save, I ask for what settled.
Have you noticed how your AI writes the notes it saves? I m curious what others have seen.

Uncomfortable pattern I keep seeing: the products with the worst UX debt are usually the ones that found product-market fit fastest. They ignored "best practices," shipped ugly-but-functional features, and got users anyway.

Then they scale and everything breaks. That's when they hire me.