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

> As a web developer, I am thinking again about my experience with the mobile web on the day after the storm

In some villages, where plenty of stone is available, people used it for everything - roof slabs, pillars, walls, flooring, water storage bowls etc. Also, villages which had plenty of wood around, they used it for everything.

As techies, we say there is an app for everything, or there is a web-technology for everything. When you have a hammer in hand, everything looks like a nail.


We also generally have large screens, faster processors, more memory, faster internet, etc.

The only problem is, when you are alone and not looking at your phone, you tend to observe people around. But unfortunately, looking at others is seen as being a weirdo, while looking at phone is considered super normal.

Also, the reason people feel comfortable with dogs is because, you don't need to act or talk in way to impress the dog, while technically not being alone. You don't get this freedom while being with people, unless you are the boss of the gang. The lack of freedom is usually offset of by the benefit of sharing, laughs and a feeling that you have achieved your goal of impressing others.


> But unfortunately, looking at others is seen as being a weirdo

No it’s simpler than that:

1) sitting alone - you’re a weirdo

2) sitting doing nothing public, staring off into space like you’re a zen master - no, you’re a weirdo

3) blogposting how you sat alone in a public space for 30 minutes and how this is an “unbearable joy” - do I need to spell it out?

This person needs help. They are having an episode. If someone has gone so far as to have this level of emotional outburst by leaving their phone at home, there’s deeper issues to unpack.

For the rest of us, sitting down at the cafe to have a nice drink and something to eat while we look at the cars and foot traffic going by is a perfectly normal activity.


Coders will become testers.

Stagnation started around 03/2014 and downward trend started around 03/2017. Looking at dates, it doesn't seem like AI caused those trend changes.

Everyone is just looking the stage to understand what's going on. No one has a script in hand that tells you what role you are going to play on that stage. The stage is full of smoke and fog. Hardly anything is visible. The entire play is being constantly re-scripted, redesigned, new pactors jumping in from nowhere, actors suddenly growing up to be monsters and filling up entire stage.

Nope we just reading HN and getting amazed. Seriously, any serious commitment requires g9od enough visibility into future prospects. You can't simply throw half a million into a garage project and then struggle to keep up with the fast changing scene on the stage.


I think that's very informative. Agree with all that. Any tools that you recommend on the context/prompt engineering side with the goals of optimizing the interactions with agents, overall LLM cost, quality and relevance of agent's responses and work output?

yes! LLMOps tools such as Opik are great for that

That's true. The man did not go around the squirrel. They both were orbiting some point near the center of the tree trunk. Otherwise, one could say that the farther point on the Moon's surface is going around the point that is facing Earth.

How does the hand-off work among the team roles? Is it a water-fall or iterative? How does the programmer complain about something that's wrong or inconsistent in architecture or requirements?

It’s iterative, not a strict waterfall, but the constraint is “text first, code second”.

A typical loop looks like this:

I talk with Perplexity/ChatGPT to clarify the requirement and trade‑offs.

The “architect” Cursor window writes a short design note: intent, invariants, interfaces, and a checklist of concrete tasks.

The “programmer” Cursor window implements those tasks, one by one, and I run tests / small experiments.

If something feels off, I paste the diff and the behavior back to the architect and we adjust the design note, then iterate.

There is a feedback channel from “programmer” to “architect”, but it goes through me. When the programmer model runs into something that doesn’t fit (“this API doesn’t exist”, “these two modules define similar concepts in different ways”, etc.), I capture that as:

comments in the code (“this conflicts with X”), and

updates to the architecture doc (“rename Y to Z, merge these two concepts”, “deprecate this path”, etc.).

So the architect is not infallible. It gets corrected by reality: tests failing, code being awkward to write, or new edge cases showing up. The main thing the process enforces is that those corrections are written down in prose first, so future changes don’t silently drift away from whatever the system used to be.

In that sense the “programmer” complains the same way human ones do: by making the spec look obviously wrong when you try to implement it.


More fodder for LLMs? I don't think humans are going to directly consume all that text.

Most researchers publish their papers to arxiv anyway which LLMs can freely access.

Don't they already train on the shadow libraries?

And 50% of that are for showing off their oddity in their social networks. The WTF factor. Do something archaic.

The same thing's happening to books where being seen as 'a reader' is a much higher priority than being well read.

https://www.gq-magazine.co.uk/culture/article/booktok-tiktok...


Or people just enjoy things. Let’s let people enjoy things.

You're both right of course, but it does seem to be an enjoyment filtered through the social media promoter lens, which makes me a little sad. Unlike say, the enjoyment I got listening to a record (and then CD) as I examined the liner notes and insert, this go-around feels like external validation by casual (or no) acquaintances. Historically this is not as valuable and can lead to some bad outcomes...

I've never seen "Let people enjoy things." used as anything other than a thought-terminating cliché. Just because something brings someone happiness doesn't mean it is immune to criticism.

But why feel the need to criticize something that doesnt hurt anyone?

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

Search: