I’m starting a new company with few good people and an exciting idea. I appreciate all the support. More updates, summaries, feels, etc will follow.

These were not the results I expected at all.

PSA: .If you want your CLI to be "agent friendly", the first step is to make sure "--yes" actually allows the CLI to run without any user interaction. Looking at "npx shadcn@latest init --yes". Who waited until I responded to the critical "Which color would you like to use as.

What's your preferred way of isolating databases?.

Looking for a running route planner app that plans routes for me based on:.- Distance I want to go.- % grade I want for the practice.- Start/finish points or general area for practice. Know anything like that?.

You can learn a lot about an engineer by noticing how they react when someone else solves a problem that they were stuck on for a while.

Is it just me, or does it feel like every paper on AWS service claims that this service is the “fastest growing ever”? I swear I read the same thing about Aurora and S3 too.

I think coding agents appeal to the same part of our brains that loves gambling. Click a button and it will either:.- Generates good code that solves your problem.- Generates messy and buggy code.- Drops your production DB. LLM is like a box of chocolate. You never know what you.

This is the most typical coding agent response ever:. "You're right! Let me check what you're referring to.".

Reason 527 to write good error messages:.It helps the robots.

Everyone uses the same ANN benchmark. It’s standardized. It’s reproducible. But is it still relevant?. A few issues I see:.🧱 Fixed workloads.There’s no scale parameter. You can’t test how systems behave across different hardware profiles the way TPC does. 📦 Only two phases:.

So today I have to say this again:. Saying "Today I have to say this again. Do X. Thank me later" is not actually good advice. Because you learn nothing (Why X?) and you can't know when this advice doesn't apply. Please give explanation with your advice, so we can all learn.

RT @tobias_petry: So today I have to say this again:. UUIDv7 primary keys are a valid choice. They eliminated the performance problems of r….

Attempt 2, after I told it to keep the bird at the exact same size shape and location as my back and limit the tattoo to my back. I didn’t say anything about the straps.

I gave ChatGPT a photo of my back and asked to generate an image of my tattoo extended with some water and reeds. Clearly I prompted wrong. I should have explained that tattoos are generally confined to the physical body.

I really shouldn't be that hard on myself. But its amazing how much time one can waste by missing a small detail in a benchmark.

Me at 8am: Mmm. why is performance (tps) on <small env> identical to <much bigger env>?. Me, to myself, at 12:30pm:

RT @AmnonShashua: We just released two papers from AAI describing recent work on our pursuit to build a “super-intelligent" AI that can mat….

I'm using this combined task list / habit tracker / self-care app. It uses AI to recommend good self-care activities based on my usual habits / tasks. Suggested workout for today: Run benchmarks 🏃‍♀️.

RT @fujii_masao: Good place for PostgreSQL hacking outside☀

RT @jlengstorf: I wrote this comment almost a decade ago and someone who inherited this code just DMed me to let me know this is still a pr….