bhu8

5 ms·

1.
by bhu8 1mo ago
I wouldn't be very bullish. Homomorphic encryption got significantly efficient with the first few iterations, but I don't really see the necessary orders of magnitude savings coming soon. You could reduce this by some partial encr
2.
by bhu8 2mo ago
Absolutely yes, but that's how you become twitter/X famous
3.
by bhu8 3mo ago
May I plugin my own stuff? https://getviberia.com
4.
by bhu8 3mo ago
Factorio/SimCity like interface for managing multiple agents: https://getviberia.com It's like the love child of Polytopia and Conductor. As many other agent management platforms/harnesses, Viberia has been buildi
5.
by bhu8 3mo ago
I like how the original triplet's initials represent their behavior well: - Opus is OP, like OverPowered - Sonnet is SO, like your significant other (this was more meaningful in Sonnet 3.7 days) - Haiku is HA, like the reaction to a ba
6.
by bhu8 4mo ago
Just use direnv? You’ll probably need to adjust the port you are hosting the local page on, but that’s just N=mod(hash based on the worktree name) and then port=default_port+N. Tell your claude to set this up. Should do it in a single promp
7.
Show HN: Viberia – Civ/Polytopia-like command center for AI agents (BYOK/BYOS)(getviberia.com)
1 points by bhu8 4mo ago|0 comments
8.
by bhu8 4mo ago
It would unfortunately impact it. ACP uses Claude SDK and is developed by a third-party.
9.
by bhu8 5mo ago
Gpt-5.3-codex is miles better than 5.4 in that regard. It’s better at orchestration, and does the things that it said it did. Haven’t tested 5.5 yet but using 5.4 for exploration + brainstorming and handing over the findings to 5.3-codex wo
10.
by bhu8 5mo ago
Feels like an issue in their caching. First non-cached turns are sent properly but everything that is second+ turn fails.
11.
by bhu8 5mo ago
I am working on a (yet another) local app for managing multiple claude/codex/gemini sessions in a game like environment: https://getviberia.com/
12.
by bhu8 6mo ago
I think something is broken though. I got 20 nematodes in a row. It's around 1% prob.
13.
by bhu8 6mo ago
Ah thanks, I see. This was 8-9 months ago. I was starting from scratch and mainly relying on Opus/Sonnet 4. I also kept running into the Godot 3 vs 4 issue before adding specific guidance about this into CLAUDE.md
14.
by bhu8 6mo ago
Great work but why not use C# instead of GDScript? LLMs are really good at C# (and tscn files for some reason), so that solves the "LLMs suck at GDScript" problem. Also, C# can be cheaper in terms of token usage (even accounting f
15.
by bhu8 7mo ago
This is amazing. I checked some games and the blunders make me think that the LLMs are not really great at forecasting what happens if they play X on Y. Can you actually introduce that into the decision making? That is, you would: 1. Have t
16.
by bhu8 8mo ago
Very nice! I am working on something very similar at www.viberia.net My take was that it’s easier to trace who is doing what (and what the agent hierarchy looks like) when agents’ locations are fixed.
17.
Token Laundering(llemre.com)
1 points by bhu8 9mo ago|0 comments
18.
by bhu8 1y ago
Noted. Thanks!
19.
by bhu8 1y ago
> Opting to introduce them sooner will almost certainly increase the complexity of your codebase prematurely Agreed, but how else are you going to scale mostly AI written code? Relying mostly on AI agents gives you that organizational co
20.
by bhu8 1y ago
Not yet unfortunately, but I'm in the process of building one. This was my journey: I vibe-coded an Electron app and ended up with a terrible monolithic architecture, and mostly badly written code. Then, I took the app's architect
21.
by bhu8 1y ago
IMHO, jumping from Level 2 to Level 5 is a matter of: - Better structured codebases - we need hierarchical codebases with minimal depth, maximal orthogonality and reasonable width. Think microservices. - Better documentation - most code doc
22.
by bhu8 1y ago
I'm Schmidhuber neutral, but the word on the street is that he is a major asshole and sometimes impossible to work with. His research might be more solid than the Turing award winners but his personality truly kept him behind.
23.
by bhu8 1y ago
I have been thinking about the exact same problem for a while and was literally hours away from publishing a blogpost on the subject. +100 on the footnote: > agents or workflows? Workflows. Workflows, all the way. The agents can start us
24.
by bhu8 1y ago
I'm sold. How do I play?
25.
by bhu8 2y ago
Also available on claude.ai
26.
by bhu8 2y ago
Yeah, the more pages I read, the more disappointed I became. Here is the reason they cite for the low performance (which is even more worrying): "The model often attempts to use a hallucinated bash tool rather than python despite const
27.
by bhu8 5y ago
Nice thread on Giorgio Parisi by Montanari: https://twitter.com/Andrea__M/status/1445405295811960841?s=2...