fazkan

I also asked perplexity to analyze the data in all of these threads, this is what it came up with

https://docs.google.com/spreadsheets/d/1_ASxAG38uP3_vcVqVnhP...

You should have guardrails, that catches these critical paths.

Shipping is a muscle, and every engineer knows how to ship good production code, just that we end up spending too much time, on perfecting tiny details. Just have good engineering discipline. Feature-flags, beta releases.

A good question to ask is why aren't you shipping every week, regardless of how critical your feature is.

looks really cool.

this is actually really cool

its pretty cool, though you might be trying to answer the questions to the meaning of life there.

I would simplify it a bit.

P.S. the correct url is fyikamran.com/llms.txt (you were missing an s)

More details https://substack.com/home/post/p-196485708

I am building docsalot to address this particular problem. It keeps the docs in sync with code

https://docsalot-docs.docsalot.dev/quickstart

We also have a full fledged web-client.

this is one of our customer

https://docs.backboard.io/api-reference/assistants/create

why do I feel that this post itself was written by AI

true, but sometimes basic ideas need to be iterated.

I agree, but honestly the example blog you shared also sounds like AI :). The em-dashes are not helping.

its a study on the market, purely based on empirical observation. Not a theoretical limit on the size of the models.

time to use claude code to understand DEKs paper, in plain English. As someone who did a bit of formal verification in grad school. I feel like, there are a long tail of problems that can be solved by human-model collab like this one. The problems may not mean much but hopefully it can stack up understanding of intelligence.

The title may be misleading, reading "diamond age" from neal stephenson right now :). Inspired from that.

Hey everyone,

I actually spent some time on it, researching the best practices, and tried hard to make it a useful reference for the future, than just another marketing blog.

Feedback on any improvement of the technical limitation, and just the language would be appreciated .

Also would love to understand anything that people do thats different than whats covered in the blog.

where did you migrate away to?

@dang now that he has made his point, shouldn't the thread be made undead?

I am sure you meant well, but it was your word-choice. Show some grace to the founders trying to build stuff. Its easy to nit-pick ideas and execution. You could have ask the same thing politely.

Focus on local, non-western LLMs, is not a unique differentiation, I am sure Peec can implement it tomorrow, if you guys want to.

Or were you only looking for differentiation so you can build it at Peec.

If two products are solving the same problem, eventually they will converge on the same feature-set. You should know that better.

What does Peec AI, do differently than Profound, Athena and a 100+ other tools in this space?

I guess, we can all be a little supportive than putting our competitors down, the market is too big

congrats on the launch, lots of competition in this space. (leap.new, replit etc). Even convex has their own app-builder.

do you guys have any blog posts technical releases, around LLM security?

just released https://changeloghub.com,

it automatically creates changelog pages from github PR commits. Still trying to nail down the PR diff to changelog generation, mostly context issues. Rn, I basically summarize each diff and then give that as a context for AI to generate structured changelogs. Works well for smaller diffs, larger diffs loose context.

its free, feedback would be appreciated

video showing how it works

https://youtu.be/bntgkZGncZ4

I will post this in defence of speed.

https://jsomers.net/blog/speed-matters

Can someone help me understand the value of this? I've build a vibe-coding product, and the existing model (both claude and openai) work well enough in apply. Its all a function of prompting/context.

I think both maybe using customers very loosely :)

yup, got lost as well. Is it like a VS-code, continue alternative?

replit, from what I know, and lovable to a certain extent.

Interesting, it seems that people don't want to share their journeys anymore.

not really, lovable, v0, and bolt all are mutlipage. They connect to supabase for db and auth. Replit can spinup custom dbs on-demand, and have a full-fledged IDE.

I did my research before jumping into this space :)

I mean zapier is also calling their workflows, "agents", I remember someone ranting about it on twitter.

thanks, do ping me if you run into any issues faizank@slashml.com