I think maybe it’s a tool and it’s up to you to make use of tools to try to let more Chinese people know and convince them to believe your idea. Don’t blame a tool but make proper use of it to make a better world.

Huh?

If you know what Chinese are suffering mentally, you may understand why I say so. Criticize a model is not the smart way to against a system.

It has been available now in coding plan.

I wonder if these local model could really solve problems especially for users that aren’t experts on a given coding language. I am not sure that, more than inline auto completion and unit implementation, are these model capable of designing and composing tech specs that really work.

Yes. It’s really not a good idea to make this ban. When the US is gradually isolated in this way by its gov’s policy, the world becomes more and more dangerous. What worse, the traditional value of open to competition that Americans have hold for centuries seems to be substituted step by step. It’s absolutely a tragedy.

People always exaggerate the thing they don’t understand.

Sometimes they exaggerate the things they understand.

Yes. But they are minority.

was referencing Anthropic

I knew it at once after you said. But politicians don’t fully understand what they face.

I hope so. But how? Who gonna fund these projects and how to coordinate with every sides. This is complex. I only believe that the open source AI won’t lack users.

I agree. Mcp might be useless in a personal scenario but it absolutely plays a role of service infrastructure in organizations. It is another form of api for those abilities that are not wrapped with rest api yet. But when they are wrapped in mcp, it seems not necessary to wrap them into rest api or cli again in near future. So these mcp services survive. The only thing matters is how to import these mcp services into agent context on demand or say by the gradual disclosure principle.

Unless you also want humans to also interact with your tools.

That’s covered in the article: a human can modify the commands generated by the agent, or vice versa, to debug problems or transfer knowledge.

This, IMO, is another scenario. MCP is designed and played as a part of the automatic tool chains. These are two different types of needs. But in the case you mentioned, when some parts of the work should be automated, it’s also possible to utilize mcp there.

Maybe the same type. Each time I call the LLM api the fan starts to work and make big noise. The temperature in the room is going up noticeably for 1-2 degrees.

> Each time I call the LLM api the fan starts to work and make big noise

So every time you do HTTP calls? Nothing there should spin up your fans, unless you use an agent with an horribly broken TUI, I've heard there is a few of those out there. But remotely calling LLM APIs really shouldn't be taxing on your local device, something somewhere is wrong/bad if that's what you're seeing.

If the horribly broken TUI you mentioned is OpenCode, I’d say yes. That’s exactly what I am experiencing.

Sure, if that's what you're using, then that's definitively buggy, unless it's doing compilation or something actually using your resources, just making HTTP calls shouldn't be heavy for your computer. Claude Code was mainly what I was thinking about, as it similarly broken, but I'm sure there are more out there as most of them seem vibe-coded at best.

Yes. Whatever *code, the same when they are working. The node.js backend is awful.

Is it a local LLM? Sibling seems to be assuming remote, but I have trouble imagining a TUI that inefficient.

No. Simply the rest api call in opencode tui. I don’t know why maybe the mbp is too old, at least it served 6 years +.

Not counting compilation passes, the rest of OpenCode is trivial enough that it should work on a 1980s PC.

It’s just so weird.

Understandable. You don’t want to lose control to your codebase and don’t trust LLM is competent in handling that fully.

No. Because they still hallucinate at times. Confuse things. Forget things. Or none of the above, as it is anthropomorphizing, but the result is the same. They can make incredible working one shots, you start to trust them, then you trust too much and .. feel the result.

Yes. I am fighting with the disobeyance of LLM on working through my pipeline commands. I believe these violations are caused by its hallucinations. So I am still developing a mechanical system to monitor agents’ behaviors automatically. I believe these routines and monitors will play as a set of scaffold to keep leading the LLM on the right way all the time.

The percentage of times I prompt claude "what about checking if there are any child processes running?" or "Would using a lock here greatly simplify the design?" only to have myself be correct is approaching 100%. That is it isn't just claude sycophantically agreeing with me. The code itself becomes smaller, simpler, and more reliable with fewer bugs.

The agents tend to produce working code but the larger the scope the bigger the mess they tend to make. They will happily evolve toward a local maxima but leave world-destroying bugs lurking in the implementation.

The other issue is that claude regularly ignores explicit instructions in CLAUDE.md or in prompts. It will "helpfully" decide to just start doing whatever it wants or reinterpret instructions completely differently than it did the last 100 times.

It has nothing to do with losing control or trust. LLMs are not conscious. They have no executive function. They aren't even thinking. They're just models predicting the next word in the script. They are very useful tools but that's all they are: tools.

I also feel like we still need to steer Claude. It doesn't always help to have stuff in the CLAUDE.md (even when it's lean). I have a lot of cases where I still need to remind the agent to do something even if it's routine.

To me I think that connects with working longer on the planning and specs. It requires reading and re-reading, but when that's done, implementation is usually much cleaner and adheres to your standards

Yes. They are tools. So my approach, at least try to approach is to keep on polishing the skills and check the output of LLM in loops with mcp to alert the abnormality asap so the LLM won’t go to next step to make things worse.

Seems have to make a face to face appointment, without any online devices in hand.