Diego Alzate

🇨🇴 trying to build cool stuff

[Diego Alzate @diegoalzate00]

https://t.co/4V4eJs3f26

[Mitchell Hashimoto @mitchellh]

I've got an agent in a loop optimizing a renderer with the goal to minimize frame times (and tests to measure). It got times down from 88ms to 2ms and allocations down from ~150K to 500. Sounds good, right? Wrong. This is exactly why agent psychosis is a big fucking problem.

As an experiment, I rewrote the Ghostty core render state in Go, with access to identically laid out data structures as Ghostty and the exact same validation tests. I made a purposely naive renderer (simple, correct, but slow). 88ms per frame with 150,000 allocations (horrendous, lol)!

I then kickstarted a Ralph loop to bring the frame times down. I told it it can't modify input data structures or the public API or tests (they're correct), but it can do anything else it wants. It got to work.

It has worked for about 4 hours. I've spent around $350 on this experiment so far. The results?

88ms => 1.5ms
150K allocs => ~500 allocs

Incredible right? Nope.

My hand-written renderer I ported has frame times (same benchmark) of ~20us (0.020ms) and 0 allocations in the update path.

This is the problem with psychosis and lacking systems understanding. If you don't understand the system, you're going to accept that this is an incredible result. If you understand the system, you'll see better solutions immediately and can do roughly 75x better on throughput.

The people who blindly trust agent output are in the former camp. They're sheeple, overdrinking from a fountain of mediocrity.

Standard disclaimer: I use AI all the time. I like AI. The point I'm making is to not blindly accept results. Think. Analyze. Learn.

[Sam Altman @sama]

today we launched ChatGPT. try talking with it here:

https://t.co/uWra8LKFMN

[Diego Alzate @diegoalzate00]

@jarredsumner does mtls work on bun?

I see a lot of open issues on this and we can't adopt it where I work.

example with reproducible case: https://t.co/bAhmOAHon2

[bashbunni @sudobunni]

Fear mongering about devs losing their jobs to AI is so boring. Show the cool stuff you're building with it if you think it's goated

no need to induce stress in fellow humans, we're all just doing our best

[Michael Truell @mntruell]

Watch Cursor build a 3M+ line browser in a week

[Diego Alzate @diegoalzate00]

@dev_TEMITAYO B, git is way easier with B

[Diego Alzate @diegoalzate00]

@stavrosnonis @vpavlin I want to see a demo!

[isaiah @tlop444]

fuck it welcome back frank ocean

[Diego Alzate @diegoalzate00]

@SCBuergel @hoprcoreandmore @hoprnet @GnosisDAO @dappcon_berlin Congrats Seb and team! Really nice work

[Diego Alzate @diegoalzate00]

Please @ChatGPTapp this is still an issue, it’s also the same experience in macOS desktop app

[Garry Tan @garrytan]

To the person who owns the Projects UI

Please un-cripple ChatGPT on mobile so that I can actually select which model is being used when I am in a project

Without model selection this UI is useless

Begging all AI agent UX people to not use Google-isms like crippled mobile UX

[Diego Alzate @diegoalzate00]

How is something like this maintainable?

@pytest.mark.parametrize("palindrome", [
"",
"a",
"Never odd or even",
])
def test_is_palindrome(palindrome):
assert is_palindrome(palindrome)

IMO this is hard to read

Are there other alternatives on the market?

[Kent C. Dodds 🐨 @kentcdodds]

Question... MCP servers that you run locally are cool and all, but I'm a lot more interested in production deployed MCP servers that are able to act on behalf of a specific user in tandem with AI tools like Claude Desktop or ChatGPT.

Are there any examples of production deployed MCP servers that users don't have to install and run themselves, which also support user authentication?

All the examples I see so far have the user generate a token and then put that token in the MCP configuration. Is that the best we have right now?

I was hoping to find an example of an MCP client that supported an OAuth flow with a hosted MCP server.

[terminal @terminaldotshop]

day 1

an opportunity to win free coffee for life

more information below

[Diego Alzate @diegoalzate00]

@zmzlois Casey is the goat, I loved his video on doing hard things

[Diego Alzate @diegoalzate00]

@stavrosnonis I don't actually use it but it exists:
https://t.co/lmptGwXEK7

[Diego Alzate @diegoalzate00]

@stavrosnonis King 👑

[Diego Alzate @diegoalzate00]

@nathudgens I agree, tooling is confusing but maybe
@astral_sh ‘s uv could help with this

[Diego Alzate @diegoalzate00]

@dhaiwat Excited to see what you cook up, chef!

[Hull City @HullCity]

➕3️⃣

#hcafc