Do people understand how bad these AIslop websites look?

I found it very tiring to read, actually.

	▲	joefourier 2 hours ago \| parent [-]
		I can get past the AI slop design myself, but the writing style is absolutely grating, especially the short repetitive phrases. "Punchy phrase. No X." "An open engine. Room to make it your own. "Same scene. Same instant. The same explosion. A different atmosphere." "A few things to know. The short answers." Why is it always the same? What awful post-training, or synthetic data training pipeline, or RLHF gone wrong, produced this homogenised, low variety slop style? And it's totally possible for LLMs to write actually decent or at least varied prose and I would think Anthropic (or OpenAI) has the resources to fix the issue, they just choose not to. Marketing copy wasn't this homogenous prior to LLMs, unless my memory fails me.

▲

jasondigitized 2 hours ago | parent | prev | next [-]

What exactly looks bad about it? These models have been trained on high converting UX patterns. They are definitely homogenous but I am not sure they look bad.

	▲	DaveChurchill 2 hours ago \| parent \| next [-]
		The homogenization of the internet. The worst corporate speak PR pages are now what gets generated for every pet project
	▲	ThrowawayR2 an hour ago \| parent \| prev [-]
		> "high converting UX patterns" That _is_ the problem: ostensibly serious writing laced with marketing / clickbaity patterns is off-putting.

▲

rfgplk 2 hours ago | parent | prev [-]

My biggest complaint (as always) is that the performance is abysmal. I wasn't going to click on it, but after reading this comment went in and tried to play the demo.

A) I was hit by a multiple _second_ load time (on a 1Gbps pipe btw)

B) The in game cursor was _visibly_ lagging

C) The gameplay itself was stuttery (and I am almost certain there is a hard one frame delay between most actions)

Then I thought, alright it could be Firefox, so I loaded it up in a fresh chromium image, and somehow it's worse. Then finally I looked at the source and it's very obvious why the performance is atrocious (the guy didn't really bother to tell his agent, "hey man, let's benchmark this and keep performance in check, maybe don't spam a billion draw calls everywhere y'know, maybe

Values below are medians across runs, including the median of each run's maximum tick time. Each pair reads *Go 1.25.0 → Go 1.27.1*; timings are ms.

|---|---:|---:|---:|---:|---:|

| Modern | 1.877 → 1.812 | 1.744 → 1.665 | 2.765 → 2.705 | 8.934 → 8.740 | 86,428 → 86,428 |

| Strict 3.1 | 1.210 → 1.101 | 1.144 → 1.041 | 1.577 → 1.498 | 4.555 → 4.594 | 19,871 → 19,871 |

| Community 3.9 | 1.591 → 1.447 | 1.433 → 1.290 | 2.407 → 2.299 | 9.323 → 9.310 | 16,721 → 16,721 |

fix this y'know ")