There are pretty hidden assumption in this comment. First of all, not every business in the AI space is _training_ models, and the difference between training and inference is massive - i.e. most businesses can easily afford inference, perhaps depending on model, but they definitely can.

Another several unfounded claims were made here, but I just wanted to say LLMs with MCP are definitely good enough for almost every use case you can come up with as long as you can provide them with high quality context. LLMs are absolutely the future and they will take over massive parts of our workflow in many industries. Try MCP for yourself and see. There's just no going back.

▲

ramoz 17 hours ago | parent | next [-]

LLMs with tools*

MCP isn’t inherently special. A Claude Code with Bash() tool can do nearly anything a MCP server will give you - much more efficiently.

Computer Use agents are here and are only going to get better.

The conversation shouldn’t be about LLMs any longer. Providers will be providing agents.

▲

anthonypasq 14 hours ago | parent [-]

correct and companies will be exposing their data via mcp instead of standard rest apis.

▲

ramoz 13 hours ago | parent [-]

That makes no sense. MCP at best is a protocol transpilation at runtime. It is not redefining things like DB drivers or connections. And I did not say rest apis enable agents. Computer use tooling does. APIs and everything else that already exists.

MCP is more like graphql. Not a new network paradigm.

The design of MCP right now is not very optimal esp when you can equip an agent with one tool vs 5-20 that bloat it's reasoning every prompt.

▲

anthonypasq 13 hours ago | parent [-]

why would you make an agent click around a web browser like a human when it could self discover the api and call it directly?

	▲	ramoz 13 hours ago \| parent [-]
		self discovery via primitives is what works well today. I never discouraged that, only discouraged MCP sensationalism. However, an agent that can see the screen and immediately click through whatever desired UI modality is immensely more efficient than swimming through protocols. There is at least one frontier lab who has prepared enough foresight that agents running on VDI infrastructure is a major coming wave.

▲

dontlikeyoueith 17 hours ago | parent | prev [-]

> I just wanted to say LLMs with MCP are definitely good enough for almost every use case you can come up with as long as you can provide them with high quality context.

This just shows you lack imagination.

I have a lot of use cases that they are not good enough for.