HNHacker News
TopNewBestAskShowJobs

adenta

374 karma · joined September 17, 2015

email: andrew at denta dot co

https://linkedin.com/in/adenta

submissionscomments
adenta··on Show HN: LLM plays Pokémon (open sourced)
Email me when it launches! (In profile)
adenta··on Show HN: LLM plays Pokémon (open sourced)
Yeah I think _a_ future of television might've been more apt. They should've made a season 5 to Snowpiercer.
adenta··on Show HN: LLM plays Pokémon (open sourced)
https://static.wikia.nocookie.net/loveinterest/images/5/5c/3...
adenta··on Show HN: LLM plays Pokémon (open sourced)
I tried! It didn’t work super well
adenta··on Show HN: LLM plays Pokémon (open sourced)
Got a link handy?
adenta··on Show HN: LLM plays Pokémon (open sourced)
Idk what the parent comment said but let’s make AI robots fight each other gladiator style.
adenta··on Show HN: LLM plays Pokémon (open sourced)
I disagree. Getting a computer to play a game like a human has an incredibly broad range of applications. Imagine a system like this that is on autopilot, but can get suggestions from a twitch chat, nudging its behavior in a specific direction. Two such systems could be run by two teams, and they could do a weekly battle.

This isn’t an exercise in AI, it’s an exercise in TV production IMO.

adenta··on Show HN: LLM plays Pokémon (open sourced)
Yeah, took a similar approach at https://github.com/adenta/fire_red_agent/blob/main/app/servi...
adenta··on Show HN: LLM plays Pokémon (open sourced)
On this page (https://excalidraw.com/#json=WrM9ViixPu2je5cVJZGCe,no_UoONhF...) linked from their twitch, it says: “ This info is all parsed directly from the RAM of the game, Claude Code is very good at this task”. I’m reading that as “we are pumping the RAM directly into the LLM”, but I could be mistaken.
adenta··on Searchcode.com’s SQLite database is probably 6 terabytes bigger than yours
That’ll teach me to post first thing in the morning.

Something something, Cunningham's Law.

adenta··on Searchcode.com's SQLite database is probably 6 terabytes bigger than yours
https://fractaledmind.github.io/images/railsworld-2024/111.p...

https://fractaledmind.github.io/2024/10/16/sqlite-supercharg...

adenta··on Searchcode.com’s SQLite database is probably 6 terabytes bigger than yours
Do you really need concurrency? My understanding is read are like… picoseconds, because everything happens in memory. You don’t have a separate server to call.

This guy is great: https://fractaledmind.github.io/2024/04/15/sqlite-on-rails-t...

adenta··on Tiny JITs for a Faster FFI
Is anyone using LiveView in production?
adenta··on Show HN: Marksmith – a GitHub-style Markdown editor for Ruby on Rails
Another question:

Does this integrate with action_text? Wasn’t that a big thing in rails 6/7, with like… trix?

adenta··on Show HN: Marksmith – a GitHub-style Markdown editor for Ruby on Rails
If I was to build out a react integration, would you recommend:

A. Rebuild the stimulus portions in react B. Write a react wrapper around the stimulus

I use rails for everything I do, but like to use vite_rails on the front end for react. Excited to use this library!

adenta··on How to Run DeepSeek R1 671B Locally on a $2000 EPYC Server
If you google, there is a $6k setup for the non-quantized version running like 3-4 tps.
adenta··on Render raises $80M in Series C financing
I’ve been pretty vocal about using http://kamal-deploy.org/ more and more, and Render less.

Kamal really is fantastic, PaaS niceties w/o the PaaS tax.

adenta··on Reverse Engineering Bambu Connect
They are locking down their software so you have to use it
adenta··on Reverse Engineering Bambu Connect
The A1 mini was my first printer and it just works.

Is there another brand that is idiot proof?

adenta··on In-Depth Ruby Concurrency: Navigating the Ruby Concurrency Landscape
I wish it was easy to use puma for almost everything, and then iodine/falcon for web sockets. Haven’t figured out a sensible solution yet.
adenta··on In-Depth Ruby Concurrency: Navigating the Ruby Concurrency Landscape
Yeah, Puma just seems more stable, and works out of the box with Kamal.
adenta··on In-Depth Ruby Concurrency: Navigating the Ruby Concurrency Landscape
This was a great talk.

The Falcon web server scares me. I wish Puma had better support for web sockets/fibers.

adenta··on Stimulus-use: collection of composable behaviors for Stimulus Controllers
As a die hard rails developer, can someone sell me on Stimulus?

I'm just now exploring turbo frames and I don't quite get the benefit. I use rails with vite. API endpoints are mounted under `/api/v1`, and frontend routes are all funneled through react-router.

adenta··on Show HN: SplitScreen – Create dual-video content for TikTok, YouTube, Reels
this should be an API. Closest thing I’ve found is IMG.LY, and it’s kinda meh.
adenta··on Show HN: Porter Cloud – PaaS with an eject button
Is there a mailing list I can sign up for to know when this drops?

Excited to try out the product! Render was a better heroku, but porter can do all sorts of cool HIPPA stuff that render charges $500/mo. for

adenta··on Show HN: Porter Cloud – PaaS with an eject button
talk to me about GPU’s? I saw some gpu_node config stuff in your documentation a couple days ago.

If Porter can host GPU’s, that’s a superpower render.com doesn’t have.

adenta··on Show HN: Sonauto – A more controllable AI music creator
I cant tell, will this let me upload an instrumental track and change the genre/instrument makeup? When I tried, I might've overwritten the prompt.
adenta··on [dead]
Is this kind of content interesting to HN readers? Seems like an ad to me.
adenta··on The Final Nail in Psychiatry's Antidepressant Coffin
I had to try five different SSRI/SNRI’s before I found one that worked. I bet I could determine my SNRI vs a placebo in a double blind trial.
adenta··on Ask HN: Examples of high-value specialised consulting?
After selling my YC W20 startup, I carved out a niche of building MVP’s for other startups.

I was worried I wouldn’t be able to differentiate myself from race-to-the-bottom dev shops, but I’ve been making it work.

I have a ton of relevant experience around helping companies launch products 2-3 months ahead of schedule, which is really important when starting something new. Often times you only have a vague idea what your customers want, so you need to move quickly & validate what they need.

← PreviousPage 3 of 6Next →