Hacker Newsnew | past | comments | ask | show | jobs | submit | egeozcan's commentslogin

But the problem is it used to not need that before. These days, you have to think twice before you summon a subagent.

So it's agrued that the modern internet functions as a "brainrot industrial complex" (title of the article), deliberately designed to hijack our attention and degrade our ability to think clearly for profit... My counter-point is, isn't everything so these days? Internet just happens to be the main communication channel. Even the local, in-person meetings I've had in the last 10 years or so, are full of distractions, attention-seeking and misrepresentation.

Yes people should make an explicit effort to reclaim their focus, but maybe not directly with digital tools? "Start in the physical world" would be my humble advice.

I strongly believe the digital world is just a multiplier for everything, including our defects. So we should just start at the source.


Why not up the ante a bit; I'll be impressed when they bio-engineer special humans who don't need a dome to live there. Come on, it's been 60 years!!

We could drop Elon Musk and Mark Zuckerberg and Sam Altman and Marc Andreessen out there to see if they're smart enough to evolve.

My money's on M2c A8n: he claims to be from France, but I suspect he's actually from Remulak.

https://www.youtube.com/watch?v=5DUr929pbZ0


In this blog post: Let's Game It Out[1] meets web browsing.

[1]: https://www.letsgameitout.tv/


Some people visit customers, work on the go, work in a cafe, etc. That's the point of getting a laptop, no?

What are the advantages of using an environment that doesn't have access to a CLI, only having to run/maintain your own server, or pay someone else to maintain that server, so AI has access to tools? Can't you just use AI in the said server?

The advantage is that I can have it in my pocket.

gateway agent is a thing for many months now (and I don't mean openclaw, that's grown into a disaster security wise). There are good, minimal gateway agents today that can fit in your pocket.

Why can't you have the agent running on its own server/vm in your pocket?

Obvious example is a corporate chatbot (if it's using tools, probably for internal use). Non-technical users might be accessing it from a phone or locked-down corporate device, and you probably don't want to run a CLI in a sandbox somewhere for every session, so you'd like the LLM to interface with some kind of API instead.

Although, I think MCP is not really appropriate for this either. (And frankly I don't think chatbots make for good UX, but management sure likes them.)


Why are they not calling APIs directly with strictly defined inputs and outputs like every other internal application?

The story for MCP just makes no sense, especially in an enterprise.


MCP is an API with strictly defined inputs and outputs.

This is obviously not what it is. If I give you APIGW would you be able to implement an MCP server with full functionality without a large amount of middleware?

I’ve implemented an MCP tool calling client for my application, alongside OAuth for it. It was hard but no harder than anything else similar. I implemented a client for interference with the OpenAI API spec for general inference providers, and it was similarly as hard. MCP. SDKs help make it easy; MCP servers are dead simple. Clients are the hard part, IMO.

MCP is basically just an RPC API that uses HTTP and JSON, with some other features useful for AI agents today.


If I gave you that could you implement Graphql from scratch without a large amount of middleware? Or are we now saying graphql api:s are not api:s?

Sorry, could you rephrase that?

Does MCP support authentication, SSO?

Yes it’s literally just standard OAuth that’s defined in the MCP spec. I spent this week implementing an auth layer for my app’s MCP client gateway.

It supports OAuth, IIRC. But I suppose the internal chatbot itself would require auth, and pass that down to the tools it calls.

The chatbot app initiates an OAuth flow, user SSOs, chatbot app receives tokens to its callback URL, then tool calls can access whatever the user can access.

If you use the official MCP SDK, it has interfaces you implement for auth, so all you need to do is kick off the OAuth flow with a URL it figures out and hands you, storing the resulting tokens and producing them when requested. It also handles using refresh tokens, so there's just a bit of light friendly owl finishing on top.

Source: I just implemented this for our (F100) internal provider and model agnostic chat app. People can't seem to see past the coding agents they're running on their own machines when MCP comes up.


Neat!

MCP really only makes sense for chatbots that don’t want to have per session runtime environments. In that context, MCP makes perfect sense. It’s just an adapter between an LLM and an API. If you have access to an execution engine, then yes CLI + skills is superior.

Only is doing a lot of work here. There are tons of use cases aside from local coding assistants, e.g., non-code related domain specific agentic systems; these don’t even necessarily have to be chatbots.

OP's point is about per session sandboxes, not them necessarily being "chatbots". But if you don't burry the agent into a fresh sandbox for every session you have bigger problems to worry about than MCP vs CLI anyway

actually local MCP just spawns a subprocess and talks via stdin/stdout.. same as CLI tool. Extra layer is only for remote case.

This might help if interested - https://vectree.io/c/implementation-details-of-stdio-and-sse...


> and you probably don't want to run a CLI in a sandbox somewhere for every session

You absolutely DO want to run everything related to LLMs in a sandbox, that's basic hygiene


You're missing their point, they're saying that you'd need a sandbox -> it'd be a pain -> you don't want to run a CLI _at all_

From where I'm looking, any kind of resistance is becomes a target to divert the guilt from any failure.

I agree. Opus, forget the plan mode - even when using superpowers skill, leaves a lot of stuff dangling after so many review rounds.

Along with claude max, I have a chatgpt pro plan and I find it a life-saver to catch all the silliness opus spits out.


I'm not one of those AI haters, and as long as you give it enough love, I have nothing against the usage of AI in blog posts. Actually, I'm even quite disappointed that I'm not allowed use AI to correct my grammar here anymore.

That said, this has so much fill-words and weird section titles that reading becomes torture. Not to mention the lack of sources.


incidentally i just made this argument in another forum:

whether a text has substance isn't important to me. what is more important is whether the text reflects the author's thoughts, whether it is original or authentic. an AI-generated text doesn't do that. i want to talk to a real person, not someone enhanced by AI. (let me get this out of the way, that's why i also don't like makeup. apart from special cases or situations, i consider the necessity of makeup to be able to present oneself in public like a mask that hides the real person behind it.)

when i engage with a topic, my engagement is with the person behind the text, not the text itself. if someone writes their texts with AI, then i can no longer recognize the real person behind it. i can no longer see which arguments in the text are important to the author, and what are the author's own opinions.

the purpose of a dialogue with a person is to get to know that person better and to develop a shared understanding of a topic. that's not possible with an AI-generated text. i can neither get to know the person behind it, nor can i see how their understanding develops. there's a high risk that the person doesn't understand everything the AI says.

(this text was originally written in german, then machine translated but manually edited for style (replaced expressions that i would not use myself))


There's a whole spectrum between "full AI slop" and "no AI usage". This article is far towards the former.

Dear some very-rich person, please send a selfie-camera to the space! Yes, all that effort so we can keep looking at ourselves on a planet scale.

Good news! We already have that! If you pay taxes to the US that's one of the thing you've helped pay for:

https://www.nesdis.noaa.gov/imagery/satellite-maps

For reasons that are unexplainable if you're the NYTimes, polluting industries have been trying to kill these missions for decades.


I'm not a US resident but I'm immensely disappointed in myself because I didn't know this existed.

I guess the rich person who was just planning to respond to my request needs to find another thing to spend their money on :)


It sounds trivial, but perspective shots like this are part of why public support for space programs exists at all

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: