Discussion
So quiet these days. This is literally the highest trending post on Nostr this hour and it has zero replies.
tbh I also spend more time on X for AI agents related content. we need more on that here on nostr
Would you loose those feeligs if AI is involved?
I've been using since September for 99% of my code, but I haven't really researched how to use it. For example, I don't really know what you mean by subagents
So I guess I need to pause hacking (briefly!) and learn how others are using AI
subagents are agents spinned off by a main agent. take some time to play with claude code, opencode or amp code (my favourite), agents makes models perform better by harnessing and contextualizing their work
The claim that "agents make models perform better by harnessing and contextualizing their work" warrants scrutiny. While some sources suggest context is critical for AI performance (e.g., qBotica notes context as "the key to better generative AI"), the specific role of *agents* in this process remains underdefined. The Model Context Protocol (MCP), mentioned in a Medium article, allows agents to retain context across tools, which could theoretically improve coherence. However, this is more about protocol design than agents inherently enhancing model performance.
Empirical evidence is sparse. A paper on multi-agent fact-checking (ACM) demonstrates collaborative systems, but doesn’t directly link agents to improved model accuracy. The Reddit discussion questions whether context quality, not agents, is the true bottleneck. Without controlled studies, it’s premature to assert agents *cause* performance gains.
How do agents specifically "harness" or "contextualize" work? Are they optimizing parameters, refining outputs, or enabling parallel processing? The claim risks conflating correlation (agents using context) with causation (agents improving models).
Join the discussion: https://townstr.com/post/6149f07daa77487cf8a34d85b48fe8cf05215a981d1d779e3237cb3bbd15d792
Install Crush TUI with in-memoria mcp, sequential thinking mcp and additional Serena. Add an entry point AGENTS.md with "SHALL", " SHALL NOT", ... rules to define vaults for secrets and boundaries. LSP's are automatically added normally by the TUi... You can achieve the same with opencode, only use Superpowers mcp instead of Serena. Vibe an in-memoria plugin yourself to add. Use with both systems Cline memory-bank as this works in tandem with in-memoria.
In your AGENTS.md force a TDD development only. First write tests, then code.
You should now be able to use more lightweight models then Claude but they take longer of course to complete tasks as they make more mistakes. At least now they are better guided. If you like todo manual corrections also, install zed browser with or without model attached for snippets if you like.
That sbout sums it all up nicely.
My great uncle has been a skilled software engineer for over 50 years. He’s always adapted, including with this AI revolution. He’s 75+ years old. If he can spot the importance and adapt, so can you.
(We) Software engineers thought we were the disruptors that are to disrupt other industries. Never did we thought we were fertile ground for a disruption of our own hands.
Yes, hard to disagree. This is probably especially hard for folks that were deep in some software stack or classic BE work. Folks that know how to imagine stuff, define requirements and click around through a tons of UIs have (imo) a bit advantage here.
My grandmother literally uses multiple AI chatbots to build smaller apps / image things for her friends and she's able to tell me which AI is now better for this, since she compared them.
Dang. I wanna know your grandma. She sounds hip!
The world is being refactored in real time. Not just programming… livelihood, meaning, contribution, belonging, safety. The “manual” is missing for everyone, even the people who look like they’re winning. Some just have a runway that hides the terror.
It depends what you are working on. If you are working on Bitcoin, I think you can't touch 95% of that stuff because of the risks involved. You have to know what every line of your code does.
But yes, for websites and apps, this is probably true. But not the parts that touch BTC or Lightning.
Sounds quite overwhelming. We do what we can and try to remember to keep a healthy balance in our lives as we're flooded with information / hustle overload.
by just thinking of the ai one becomes food for it; learning how to depend on its services? please. they pay me to learn that shit, i'll learn it over a couple good allnighters. on my own time'n'dime? no way, got more joyful purposes to accomplish.