terabytest

What’s the difference between a law and an integration test?

Sounds like the definition of monarchy?

I apologize if this came across as an attack. I would just say that in a sea of slop, it’s really hard to invest the time it takes to judge the quality of an individual tool, and you might develop heuristics to survive. One useful heuristic is whether the documentation looks written by AI or not, because the level of attention paid to documentation often mirrors what you’d find in terms of software quality. If you want to stand out, you have to look like you’ve invested the time and effort to communicate to your reader that what you wrote is worth their precious time.

At this point, are we sharing these AI-written articles to make fun of how bad they are? Is there any substance left to discuss?

> When the panel proposes a change to the file, the Änderungen prüfen modal opens first.

This is in the first section. If they never bothered to proofread their article, how is it worth the time it takes to read it?

I’m not being facetious. It’s becoming extremely hard to separate the wheat from the chaff. Every piece of software looks and reads like a homogenous pile of Claudisms. Even if the software in question is good, I immediately lose trust in the developer’s good faith when I see stuff like this.

For what it's worth, I would love to read the earlier drafts before the AI workshopped them.

To be clear I’m not making any statement or implication about Ryan’s work or his quality as a developer. As someone who knows nothing about Solid I’m trying to read the article and understand what the USP is and struggling. And remarking on the AI tells in the writing.

As much as I’m trying to understand what’s groundbreaking here, and I’m really trying to appreciate it, the prose style of the article makes it completely impenetrable. I’m trying to understand what makes Solid 2.0 special, but I’m coming up completely empty, because it feels like I’m just reading noise.

Out of curiosity, did you use AI for this reply too? It reads a bit like it.

> I won’t be using AI in this way to create any more web stuff, and I do not use it to anything close to this extent on iOS software. I do ask questions, debug issues, things like that, but I do not create apps with Claude for iOS.

I’m worried this is another lie. “Current,” Terry’s other app, is very obviously built by AI. Its interface carries all the tells (and slop-induced UI bugs) that you’d run into if an app was primarily built through vibe coding. I won’t get into the merits of whether the app works or is useful. I bought it and immediately requested a refund after I noticed all the AI tells.

LLMs are not as cheap on enterprise plans.

Is the issue here a lack of “Reject All” button? Or strictly the number of partners?

This one honestly drives me nuts.

Another issue I had recently is one of the subagents made up a bunch of meaningless jargon and Claude gladly repeated it to me with no definition and completely assuming I was coming along for the ride.

As a EU citizen I’m at a loss for what to do about this. I feel that they’re going against any average citizen’s interest. What can we do to make them stop?

We truly are reaching the end times when an article criticizing the use of LLM is in itself pure AI slop.

You’re suggesting flooding the lungs in order to get a scan?

I have the feeling that this article was hurt rather than helped by being written using LLMs. It was really hard to follow, and even though I read it hoping to learn something new, I left feeling more confused than when I started. The feeling while reading was that the prose was trying to hold my hand but had absolutely no empathy for the build up of my understanding over the article. It’s a bit like when, as a child, you’d do homework with your parent and the parent would start saying “don’t you see how it’s obvious that 25/5=5” with no further explanation and a building tone of frustration.

What is the 3d printing technique being used here? I can’t intuitively recognize it.

Cool! But I'm curious why nobody's gone and published a complete app for multiple platforms? Seems like an obvious next step.

Are there any good end-user apps that are supported on a broad set of platforms and use Iroh to support file transfers between e.g. Windows and iOS seamlessly?

I recommend revisiting this comment when you have a son or daughter.

Where do you get the money from while the lawsuit is ongoing?

The atrocious writing style makes it at least obvious this wasn’t written by AI. Silver linings.

Wispr Flow is a masterclass in STT. Apple's solution feels like it's from the last century in comparison. Same applies with Apple's TTS when you have ElevenLabs and OpenAI running laps around it. All I need is for my iPhone to do those things natively at the same quality level (because in Apple's walled garden that's the only way to get them usable everywhere).

If you can, please take a look at the discussion I started on the repo. https://github.com/saffron-health/libretto/discussions/227#d...

I haven’t used libretto myself yet but I’m excited about having this kind of tool at my disposal as it’s been a need in the past.

Are they selling this anywhere or is it just a marketing gimmick?

Looks awesome, but I wonder if its functionality could be exposed to existing CLIs such as Claude Code instead of having to run it through its own CLI, mainly because I don't want to spend on credits when I've already got a CC subscription.

EDIT: To clarify, I realize there are skill files that can be used with Claude directly, but the snapshot analysis model seems to require a key. Any way to route that effort through Claude Code itself, such as for example exporting the raw snapshot to a file and instructing Claude Code to use a built-in subagent instead?

Why did you choose this plan in place of a fixed price plan?

How is this on the front page? This reads like pure AI slop. It feels like an insult to the reader.

OP: if you thought you had something useful to say, why didn’t you write it in your own words. There’s no useful content I can discern while reading this post.

> FastRender may not be a production-ready browser, but it represents over a million lines of Rust code, written in a few weeks, that can already render real web pages to a usable degree

I feel that we continue to miss the forest for the trees. Writing (or generating) a million lines of code in Rust should not count as an achievement in and of itself. What matters is whether those lines build, function as expected (especially in edge cases) and perform decently. As far as I can tell, AI has not been demonstrated to be useful yet at those three things.

That’s not what I meant. What I’m asking is whether there’s any evidence that the latest “techniques” (such as Ralph) can actually lead to high quality results both in terms of code and end product, and if so, how.