Congrats on the launch. Sign-up was super smooth and the product is compelling.

Will be testing it with some of the agent work I'm doing on agent.ai.

If you're ever looking for a super low-maintenance indie investor, please reach out. Have lots of YC folks that will vouch for me.

One of my favorite books on the actual practice of programming.

The other one is "Code Complete".

Code Complete is my favorite, hands down. The Psychology of Computer Programming was also very much worth the read.

If you are into podcasts Steve McConnell (the author) was a guest a couple of months ago on the pragmatic engineer podcast. My one take away from that podcast was just how young and new in his career he was when he wrote that book.

https://newsletter.pragmaticengineer.com/p/code-complete-wit...

Yes. And if you read his other books, you'll see that he was trying to create positive change in programming over and over again, but steadily climbed what part of the organizational ladder he was trying to address.

That book, being the first, is pretty squarely aimed at shortening the journey from junior programmer to senior.

That's fair feedback. We're working on making the UX better -- especially around the Agent Builder.

Lots of work to do, and we're still a tiny team (but growing).

Thanks for noticing my post, building agents.json and open sourcing it.

This is why the Internet is awesome.

You're the best :) thanks for inspiring

Very much clicked for me.

Loved this line:

"I’ll discuss a few boring things we do that help us keep things boring."

Looking forward to the HubSpot + MySQL integration.

Disclosure: I'm one of the HubSpot founders.

Hubspot and MySQL are both next on our list for deploying out of beta! I'll shoot you a message when they're out :)

Thanks for including a search provider for HubSpot and best wishes with the launch.

Disclosure: I'm one of the founders of HubSpot so a bit biased. :)

I'm from HubSpot (I'm one of the founders).

This was our second year doing the "Week of Rest". It was an experiment we ran last year.

It ended up being a good idea and works well.

This is a really useful thread with posts from people actually using LLMs.

I wonder if Claude-100k could be used to ingest this entire thread and then answer questions based on it, or summarize or identify the pros/cons of certain aspects of Claude, large context windows, vector embeddings, etc.

This is how I've been thinking about it.

Hybrid interfaces that combine visual cues and output and natural language input.

That's what I had in mind. Happy to talk to the computer but show me the results on a screen. If I'm typing in a freeform query, I'd generally just as soon say it instead. (Assuming I'm somewhere I can talk.)