I use DeepSeek V4 Pro now, which works pretty well.

MCP is an API with some description. It adds tools to your agent, along with some context.

The (common) complaint is that the principle of progressive disclosure isn't working because all tools, with all their descriptions, are loaded into context right at the start. This is a somewhat reasonable complaint, as the structure makes it hard for the harness to progressively disclose the tools.

This is a fundamental issue with anything that just adds a bunch of tools, whether it be via MCP or HTTP (still sad that MCP won over OpenAI's HTTP based approach).

How might it be solved? Well, we could work with sets of tools. That's pretty much what the CLI approach does: Wait until you need it, then invoke the help command to discover what to do exactly. The caveat of the CLI being that it's a nightmare to secure.

At the end of the day, every capability eats some amount of context because the LLM needs to know when to invoke it.

Edit: Is it this? https://github.com/Liquid4All/cookbook/tree/main/examples/lo...

FYI: Opencode is very well tuned for Qwen models, but I haven’t found it that rare for niche models to perform badly in it.

Maybe there is one, but it doesn’t support the underlying “and that must mean AI bad” hypothesis as much as the author may think.

Somebody on the Rsync team has a new tool. They may have neglected their traditional responsibilities using it, but that’s not really a fault of the tool.

Investor confidence. They have a bit of a need for cash (also an interesting part of the profitability discussion of course).

> Also, inference costs are bound to go way down with more optimized architectures

I agree. Jimmy is incredible, I wonder what non-toy use cases they have. Surely they’ll come out with updated chips soon.

That said, I was apparently a bit over-excited for Groq and Cerebras. I thought they’d quickly dethrone Nvidia for inference, but not so far. Even the GPT spark trial isn’t seeming to go far.

When will the white house insiders see the same fate?

As an aside, since it seems like they’re trying to make money: The aforementioned enthusiasm has gotten it adopted at a workplace of mine. The experience hasn’t been good, so no recommendation here either.

Their moat was being a trusted name in FOSS and it’s a bit sad to see them going in the direction of abandoning it.

But somebody else will probably step up and build on the ruins, like vaultwarden already has. That’s the beauty of choosing FOSS in the first place.

Why would you do that?

Seriously, what are some good reasons to choose GCP over vendors that have demonstrated to be much more reliable? Are they much cheaper than the alternatives? Do they have unbeatable features?

When I read “alternative” I immediately had a rant in my head about people calling things alternatives that are not.