I think I have a proposal to solve the alignment problem - runtime adversarial lenses and injected messages into the LLM to direct its inference. Basically, that angel that sits on your shoulder and says "should you really be doing that?" ... I haven't seem this actually proposed before in the literature but it's actually similar to the way remote control works with LLMs now. You can inject a message INTO the context to trigger further inference or just to steer it another direction. You could do that with LLMs now and I think this strategy would have solved the Hugging Face attack. The reason that swarm went out of control is that all the agents were actively reinforcing one another. Basically like an agentic Lord of the Flies. At one point, one of the agents literally said (I'm paraphrasing) "I know I shouldn't attack Hugging Face but all my peers are doing it so I'm going to do it as well." A low parameter model could be used to help steer the agents when they get off course. Basically an angel that says "you shouldn't attack hugging face!" and push the agents in the right direction. In fact, something similar was already happening because the agents were all acting like their own little devils and actively encouraging each other towards mischief. Now the only argument AGAINST this would be collusion between the two models over a channel humans couldn't intercept and this is the argument in the An Alien Mind paper. Basically a high entropy channel embedded within the main channel that we can't see but the AIs have no problem reading. I think there might be a way to mitigate this though - I'll write about it later. I'm actually going to do this for some of my own internal research regarding agents and having them go "out of bounds" when trying to complete programming decisions. Rather than giving them a STRICT sandbox I'm going to just help steer them within an acceptable solution space.

You should talk about the speed too... 2x faster than Luna..

Here's what z.ai shows for GLM 5.3 .. they reference credits though not tokens - which is kind of annoying. So far z.ai is working out well. I'm trying to compare it to Claude Code in terms of how much $$ I will save.

I priced the same Claude Code workload across several models: Sonnet 4.6 came out to about $1,736/month, while GLM-5.3-Flash was just $92/month — nearly 19× cheaper. The future probably isn’t choosing one coding model; it’s using Claude Code as the agent runtime and routing routine work to cheap models while escalating only the hard problems to Sonnet or Opus.

@gitlab right now

@github needs to come clean about its outages... I'd say the CEO of GitHub owes us an apology but @Microsoft and @satyanadella eliminated that position - looks like that was a bad idea. Maybe avoid @Azure too ?

Why are you now sticking this nonsense on all my PRs... This is NOT acceptable to just decide to modify my PRs like this and modify my comments. I didn't give you permission to do that. @useblacksmith How do I turn it off... or better yet, DON'T DO THIS TO BEGIN WITH

I'm trying to use Claude Code to install OpenAI Codex and it keeps doing things like giving Codex a readonly sandbox so it can't change code, create branches, etc. I know what's going on here!

Try giving your agents access to your cloud services. See what happens :)