jdthedisciple

Well, a weekend project that requires 345'600'000 tokens — you only get so many seconds in two days ;)

On a whim I went over to try and discover public Github gists to see if there is anything interesting or valuable–

Only to find that it is getting absolutely spammed to death by chinese-looking SEO-farming posts.

I'm surprised Github hasn't picked up or managed to prevent this. When it's this obvious.

Howcome?

It obv does have an extremely natural humane sound to it, but I feel like if you pay attention to what it actually says there is a noticeable lack of intelligence behind it.

It honestly sounds a bit like a teenager's stream of consciousness to me, that's how they sound to me anyways.

Finally a serious benchmark!

That's only bc folks forget to specify "pick svelte and make it efficient"

I 100% get the worry and yes that happens a lot.

There are easy ways to prevent the bloat and wasteful performance though, sth I do deeply care about.

You can get a long way staying on vanilla or picking (for your llm) one of those superfast frameworks like svelte in ny experience :)

Tbh if it's really crucial I let the PR sit there for a while and sleep over it, poking at it every now and then over the course of a cple days, until I'm comfy w it ...

Exactly, we don't review UI. Nobody cares as long as it looks right and feels smooth.

A total overhaul is always but a single prompt and 5 minutes away, should the need arise...

At the end of the day it's really about developing a feel for it:

You will know intuitively if it's worth taking a closer look or if u can safely close your eyes and just commit it right away.

I usually know which meta questions are crucial to clarify with colleagues and where to check the code punctually.

I don't get all the fuzz about AI in general:

Are we worried there won't be any problems left for humans to solve anymore?

If you think so... why don't you just go and celebrate?

Yea but you chose to blame the LLM, not the vibe coders.

As I said: prompting correctly and reviewing are still indispensable skills.

Ofc a big part of it is still knowing how to prompt correctly and manage context.

Perhaps a lot of devs still have a flawed mental model of how these things even work.

Whenever I stop for a moment to think about it:

It is absolutely batsh*t crazy insane what SWE has come to in the last cple years.

And yet part of me is loving it.