prohobo

7 ms·

1.
by prohobo 15d ago
Some anecdata: - It's extremely verbose and often incomprehensible when doing even basic tasks. Like it'll write a giant jargon-filled essay then end it by asking for a judgement call on something that references its own convolute
2.
by prohobo 1mo ago
Both of your comments are illuminating :p So, we could technically debug a prompt's output? I get that there are too many steps to actually step thru, but what if there were checkpoints? At least you could isolate behaviors to specific
3.
Offpunk Manifesto(blog.ayom.media)
2 points by prohobo 1mo ago|0 comments
4.
by prohobo 1mo ago
Well it's a transitional methodology during the early stages of a major shift in software engineering. Just look at what almost everyone is complaining about with AI coding, and then the inadequacy of their proposed solutions - this
5.
by prohobo 2mo ago
It's clear to me now that the lack of benchmarks for competence in this space means everyone thinks they're geniuses who don't need a better process, because apparently their workflow "works for them." Not sure what
6.
by prohobo 2mo ago
I agree, I tend to give a high-level concept for what I want done first, and then break it into smaller tasks. That said, there's almost always a misunderstanding within the smaller tasks anyway, which I have to test to find out about,
7.
by prohobo 2mo ago
Thanks, yeah I just went with the most controversial term as the catch-all which just means: you prompt an agent as your core workflow. For me as well, I do projects for clients that would have taken me months, and I deliver them in a fract
8.
by prohobo 2mo ago
I'm making the case that we should make tools for the higher abstraction, not a lower one. You said it yourself: why should we read/write binary? We don't, we shouldn't. It might be useful in some cases, but working wi
9.
by prohobo 2mo ago
I never said you have to turn your brain off to feel the limitations. Actually, for more complex work I think it's pretty common to spend a long time crafting some elaborate prompt, and then arguing with the agent for 10-20 turns, gett
10.
How do we stop vibe coding?(alexklos.ca)
61 points by prohobo 2mo ago|76 comments
11.
by prohobo 2mo ago
The control room in Jurassic Park was honestly beautiful - and something about those shaving cream cans really crystallized the aesthetic for me, weirdly. I had no idea that it was taken so seriously, I just remember thinking "damn, th
12.
by prohobo 2mo ago
Bringing model-driven development to coding agents with a desktop tool called Scryer: https://github.com/aklos/scryer It's still pretty experimental, but I think now that AI can feasibly generate code from intent&
13.
by prohobo 2mo ago
This is all from the perspective of people who already oppose Farage, though. Supporters will zealously reframe everything as illegitimate attack. Soft skeptics will notice the establishment's overreach. I agree this might be a normal
14.
by prohobo 2mo ago
I lived in London for six years starting just before Brexit, so I watched this exact confidence up close. The British public "knowing" Farage is damaged is not new, it's not justified, and it doesn't do anything except m
15.
by prohobo 2mo ago
This doesn't make a fool of Farage though! It clarifies to anyone already skeptical of the establishment that when pushed, they'd rather not play than risk losing. This is literally self-defeating! You claim that this kind of thin
16.
by prohobo 2mo ago
I don't understand why the mainstream is celebrating the total collapse of any meaningful political opposition to populists. This kind of fluffed-up 2010's style optimism campaign is exactly how the establishment lost control in 2
17.
by prohobo 3mo ago
RE: Tauri not having cross-compile... There's a GitHub action that compiles for Linux, Windows, and Mac. So practically it does have it, just not out of the box.
18.
by prohobo 4mo ago
Just started watching Widow's Bay, which to me has a similar vibe, but a much better hook: folk horror as the plot driver. I can't watch Northern Exposure anymore. It's too alienated from the modern world, it feels like an ar
19.
by prohobo 4mo ago
They're in every discussion even remotely related to anything Elon Musk. Everything Elon does is somehow stupid or evil. Actually, that reminds me of Thunderf00t YouTube streams where he was (or still is?) betting Starship would fail m
20.
by prohobo 5mo ago
I felt this way from a year ago up until February 2026. Claude Code and Codex becoming the norm cemented for me that a lot of the projects people are working on (including mine) are totally obsolete. As far as I'm concerned, most code
21.
by prohobo 5mo ago
I don't get it, you joined this thread to call me an idiot with a meme, and now you're talking about being a neutral arbiter for a technical discussion that I supposedly ruined. More than anything I'm getting frustrated with
22.
by prohobo 5mo ago
Just give us a taste of what we'd be paying for? I'm sure you're an expert but before I commit to 2+ years of consultation I'd like to see your approach.
23.
by prohobo 5mo ago
You're clearly very intelligent and a real software engineer, maybe you can explain where I'm wrong?
24.
by prohobo 5mo ago
This is a design problem, and not something necessarily solved by CLI --help commands. You can implement progressive disclosure in MCP as well by implementing those same help commands as tools. The MCP should not be providing thousands of t
25.
by prohobo 5mo ago
"MCP is less discoverable than a CLI" - that doesn't make any sense in terms of agent context. Once an MCP is connected the agent should have full understanding of the tools and their use, before even attempting to use them.
26.
by prohobo 5mo ago
You do understand that what it sounds like you're talking about is essentially a proto-MCP implementation right? Except more manual work involved.
27.
by prohobo 5mo ago
How would the AI know about the calendar app unless you make the text file and attach it to the session? Self-describing APIs require probing through calls, they don't tell you what you need to know before you interact with them. MCP
28.
by prohobo 5mo ago
Why would I do that if the MCP already handles it? The MCP exposes the API with those tools, it explains what the calendar app is and when to use it. Connected MCP tools are also always in the model's context, and it works for any AI a
29.
by prohobo 5mo ago
Let's say I made a calendar app that stores appointments for you. It's local, installed on your system, and the data is stored in some file in ~/.calendarapp. Now let's say you want all your Claude Code sessions to use t
30.
by prohobo 5mo ago
Right, that feels like something you'd do with a script and some API calls. MCP is more for a back and forth communication between agent and app/service, or for providing tool/API awareness during other tasks. Like MCP for

More ›