I have to say this again, and again, but it is VERY IMPORTANT when working with LLMs and making sure you don't run out of tokens. Do not ask the AI to do something only once (i.e. verify a screenshot), ask the AI to create a system/script to always automatically verify and report those issues in an optimized way. Also, once a system is implemented, don't stop there. Ask it again, after a few runs, to check the executions and see if it can improve performance/efficiency of the system. And, not only once, this improvement check should be done regularly over time, as your tasks or data do change, so the system has to be "recalibrated". For example, I just tried using Cludflare crawl/Browser Run to get some info automatically, and Astra xhigh implemented the basic API call to get the data. After this, I suggested to try again, but see if it can reduce the Browser Run time or load, while still retrieving the same data, it did it: "The light web retrieval test worked: 0.78 browser seconds, down from 4.34 seconds—about 82% less for the same product. Data accuracy checks still passed."

Am I dumb? What even is this question from Opus: "Should the vendored flipMath.js ever be re-extracted from the patched page-flip package, or can the page-flip dependency and its patch be dropped once the orphan PlacementPageTurn.tsx is removed?"

Anyone knows any snipboard.io (former snag.gy) alternatives? It seems to be gone. Websites where I can just CTRL+V images from clipboard to share them.

Wait, with Claude plans you also get $100 API credits for free per month, on top of your weekly usage??

POV: you post your new app on Reddit and start reading all the positive and constructive feedback.

GPT-6 Astra is so much better in practice than GPT-6.1 Sol. I can trust it a lot more for some reason.

Wow, all SOTA models still can't write good website copy. The seem to fully understand the context, who the target audience is, and, most importantly: what matters and what is irrelevant. I used Astra and Opus 5.5 trying to update a landing page copy, it had all the info needed but still wrote AI-like marketing copy that makes no sense to a human.

Huh? Apart from the lack of any actual numbers... So the model got more efficient, but they directly say "yeah, we made it more efficient, but we're not going to pass those savings onto you - we are reducing how many tokens you get, so all those efficiency gains don't translate in more usage". Which is ok, if they want to maintain the same offering at the same cost, but: - the older models are as efficient as they were before, so you get less usage for them - why promote efficiency as a customer benefit, if the customer is not the one benefiting from this improvement? yes, more efficiency also mean faster responses, but those are separate benefits What am I missing? Sorry for the rants, I am just trying to figure out reasons to renew my Pro subscription, and I can't really find much.

It's funny that changing a letter on my homepage nowadays goes through like 5 different powerful PCs. Local PC -> Local Dev Server -> OpenAI LLM servers -> Local Dev Server -> DNS/routing servers/switches -> git servers -> Website Hosting Dedicated server and all the way back...