alex7o

6 ms·

1.
by alex7o 6d ago
Just ask claude to port it to latest gcc :P, pls don't
2.
by alex7o 7d ago
This looms nice and I starred but you just want actors and there are other options for that you might want to compare, I like rika actors, or temporal (not actors exactly but durable state and workflows). This is a space where a few very si
3.
by alex7o 8d ago
Hacker news is still fine tho?
4.
by alex7o 8d ago
I think this is cool not for kimi but for sth like glm flash
5.
by alex7o 12d ago
You are most likely correct but that sounds less fun. Also they can do it much smarter way. Just ask these companies for content so they can detect it and prevent copyright infringement
6.
by alex7o 12d ago
I think they are training ai porn so ... this should be a lawsuit
7.
by alex7o 12d ago
I have a circuit board on my bench designed and ordered from china for 1h (manufacturing+ shipping took 14days), fable designed it I think or sol don't remember anymore but it was in claude code. I told it to benchmark different kicad
8.
by alex7o 13d ago
I think not are the same the only thing that differentiates them is adoption. If people show interest in the thing you are a genius if they don't you are mad/crazy.
9.
by alex7o 13d ago
Maybe it is AGI and they didn't benchmax it or it is not and is worse then 5.6 sol, which if true would just be sad
10.
by alex7o 13d ago
I hope they don't `fable` it and block people from doing they daily jobs with it, by introducing huge amounts of restrictions that are not really needed.
11.
by alex7o 16d ago
This sound very similar to https://www.geldata.com/ you might want to check it out.
12.
by alex7o 17d ago
My biggest problem is that msps can't be chained together and that the response schema is not know models are very good at filtering with jq to get only the relevant fields
13.
by alex7o 18d ago
I would argue that if you have the same setup as the author then use it. Yes a moving target but won't you fix something you use for yourself? Also point an agent at it get it back working in a few hours of waiting :P
14.
by alex7o 20d ago
In the non China countries. But the Chinese populace would defend chinase companies unless they do something against China themselves
15.
by alex7o 24d ago
What if, we put the sqlite driver in kernel space? So this is your only filesystem like a unikernel
16.
by alex7o 25d ago
This looks cool a chiplet can fit 30m params so the biggest card can fit qwen 3.8 27b it would be cool to see some benchmarks on things like that publically.
17.
by alex7o 27d ago
I think surrealdb is actually nice, just a bit missmanaged
18.
by alex7o 27d ago
Cant they just put ai agents at optimizing their slow internal paths.
19.
by alex7o 27d ago
Man you could have used this a small piece of marketing with a shareable link to show people how did you reason about the product.
20.
by alex7o 27d ago
Amazing, we need more DBs that can be queried from the browser directly. I needed something like this a lot but I ended up using sqlite with ranges instead
21.
by alex7o 1mo ago
use a browser that saves tabs on disk like zen browser I sometimes stop it when playing games since it will preserve tab state
22.
by alex7o 1mo ago
Taste I suppose?
23.
by alex7o 1mo ago
Oh that Is cool especially on a foldable you can run some interesting apps on it. And with wine + fex even more crazy stuff.
24.
by alex7o 1mo ago
This is a harness problem not a model problem, try prime agent it can do that and it will do it well even :P but you need to prompt it in according to its tools and processes.
25.
by alex7o 1mo ago
That is great work guys, for me turning that into an mcp for agents and assistants has been an amazing experience
26.
by alex7o 1mo ago
This was also posted 6 months ago: https://news.ycombinator.com/item?id=47155526
27.
by alex7o 1mo ago
Nah, local models with abliterated layers and uncensoring will be sold on the deepweb
28.
by alex7o 1mo ago
I see the reasoning is always last, but I have seen models do better if you put the reasoning as the first json fileld
29.
by alex7o 1mo ago
Some people might want to remove it, but onnx is one of the few ways you can run hw accelerated convolutions
30.
by alex7o 1mo ago
To be honest I am thinking of an entirly different situation, ask the llm to create Red tests for you. Then implement them yourself. You don't waste time writing tests but you still need to spec what you implement properly so the llm c

More ›