bogdan ๐
@boqdan_
Followers
877
Following
1K
Media
107
Statuses
693
creativity maxi ๐น bringing FX on-chain @MentoLabs๐งช prev @clabs, @consensys, @anyroad, @soundcloud โผsemi-professional hodler
the white lodge
Joined June 2009
80% done on an explorer for the tiered memory system in Solum, my Elixir-based @openclaw killerโข๏ธ The context is just a frame. I've been running this system for a couple of days. It is no longer stuck in the user->agent message loop, and through context engineering, it treats
38
33
341
Excited to see this hit the chains in the following weeks, after so much work.
0
0
1
could we just like move all LLMs to Europe for the summer and take a break?
0
0
3
๐ฏ I wish I had numbers on this.
Does anybody else feel like Claude Code keeps nerfing their usage limits??? I have the $100/month plan and 30 minutes of coding now uses up 60% of my session limit... A few months ago this would have only used like 5% I think. Anybody else feeling the same way?
0
0
1
to the moon ๐ฅฎ, but actually let's just go.
yacine is right. the gap between what the public thinks is happening (better autocomplete) and what we are actually shipping in the lab (autonomous reasoning loops) is wider than ever. once you see agents self-optimizing via rl in real-time, you realize the 2020 covid analogy is
0
0
3
Ah man, @AnthropicAI really doesn't want you to use your subscription for anything else. Has anybody else noticed Claude Code writing this to the credentials file on token refresh, probably? ๐
0
1
3
Iโm slowly becoming convinced that AGI might look like a high-frame-rate input model taking in multimodal sensory data, outputting contextual embeddings sampled by another model. That model stacks a self-compacting history of perceptions and outputs actions or self-reflection.
0
1
2
@openclaw
https://t.co/ZGeaonBhyF โ join the beta tester's group.
0
0
1
@mikehostetler I feel like you'll like what that first codex is working on.
0
0
1
The team is working hard. Also, fyi, I am going to obsess over the UI. Sorry, not sorry. Join beta โ
2
1
10
Been running this agent with sedimentary context compaction for the last 4 odd days on Claude, and just managed to wire it up to switch the harness fully to Gemini 3 Pro. He's so fucking rude I can't even.
1
0
1
I think around 5 years ago, I started screaming at people about local first software. Never gotten into personal hardware but damn this is the most beautiful thing I've seen today.
I built my own graphics library and got a whole map of my city onto a device with only 64kb of free RAM using my own rendering method written entirely in C & assembly, and a converter that take GeoJSON files and compresses them - this one went from 5mb to 34kb in my proprietary
1
1
75
@noself86 @jefftangx Probably the funniest part of all this is that we can now happily call them AI daemons without a second thought.
0
1
1
@openclaw For the why not OpenClaw questions:
@alexph_dev @openclaw Yeah, I struggled with that, but there are a few major things OpenClaw really lacked imo in the core loop, a few off the top of my head: JavaScript is brittle. It was made to move text around a page Erlang is basically built for this: long-running distributed systems with
0
0
4
Why did I choose to build Solum, my OpenClaw Killerโข๏ธ, instead of making OpenClaw better: (wasn't an easy choice btw)
@alexph_dev @openclaw Yeah, I struggled with that, but there are a few major things OpenClaw really lacked imo in the core loop, a few off the top of my head: JavaScript is brittle. It was made to move text around a page Erlang is basically built for this: long-running distributed systems with
0
0
11
Reply or dm GitHub handle for beta.
@jefftangx Iโve just landed an MVP a few hours ago. After 2 weeks of power using openclaw and being frustrate. I built it on elixir (erlang). GenServers are the best abstraction for long running AI agents. Proper fallback, restart, messaging, mailboxes, great concurrency primitives. Iโve
6
0
3