Hacker News: rpdillon

New comment by rpdillon in "Perlisisms (1982)"

rpdillon — Sun, 14 Jun 2026 19:07:39 +0000

My favorite has always been:

> 31. Simplicity does not precede complexity, but follows it.

Kind of close to "build the first one to throw away".

New comment by rpdillon in "Perlisisms (1982)"

rpdillon — Sun, 14 Jun 2026 19:04:47 +0000

These were published in the Communications of the ACM in the 1980s, I discovered them in the early 2000s, and have been reading them annually since. Every year, one of the ones that didn't make sense to me the previous year suddenly does.

In this particular quote, Perlis is talking about relevancy to the problem. He's hinting at the difference between incidental complexity and inherent complexity. Inherent complexity is a property of the problem, and incidental complexity is a property of the solution. He's arguing against solutions that bring incidental complexity that requires attention to aspects that aren't relevant to the problem.

New comment by rpdillon in "No, everyone is not using AI for everything"

rpdillon — Sun, 14 Jun 2026 16:50:27 +0000

> But ironically, I used AI to build it

This is a pattern I encourage - the AI might not be reliable, but with coaching, it can produce reliable tools. `colordiff` was causing issues with `less` when I was looking at diffs (character encoding issues I think), and when I asked Kimi K2.6 what to do, it built me a rust command-line diff tool in one shot that I've been using ever since (it even downloaded rust, wrote the tool, and compiled it).

New comment by rpdillon in "GLM 5.2 Is Out"

rpdillon — Sun, 14 Jun 2026 14:24:06 +0000

> effective

Depends on the task.

New comment by rpdillon in "Show HN: FablePool – pool money behind a prompt, and Fable builds it in public"

rpdillon — Sat, 13 Jun 2026 16:03:48 +0000

Using a harness to build something should clear that barrier, by the same logic that a photographer pointing on camera at something and pressing a button clears the barrier.

New comment by rpdillon in "I Am Not a Reverse Centaur"

rpdillon — Fri, 12 Jun 2026 20:00:07 +0000

It's mostly that how much you decide to involve AI as a spectrum. To extend to the pizza analogy, I feel like you're telling me that because I used dough that I bought at the store, I shouldn't be proud of the pizza I made, even though I made the sauce and cut the pepperoni and the sausage and baked it myself on a peel covered with cornmeal. That's not the same as just ordering it on DoorDash.

New comment by rpdillon in "I Am Not a Reverse Centaur"

rpdillon — Fri, 12 Jun 2026 19:55:51 +0000

But you're acting like everything that you use AI to build is easy to achieve, and that doesn't seem to be true.

New comment by rpdillon in "I Am Not a Reverse Centaur"

rpdillon — Fri, 12 Jun 2026 19:51:46 +0000

You're assuming that everybody will be equally skilled in using an LLM to create software. I don't think anything in my experience indicates that this is true.

New comment by rpdillon in "I Am Not a Reverse Centaur"

rpdillon — Fri, 12 Jun 2026 19:48:44 +0000

I think there's a spectrum between simply writing a prompt and generating slop and using AI in a loop over many hours/days/weeks to produce something that works the way you want it to. I get a great sense of accomplishment from doing the second, and I pretty much refuse to do the first, except only in the most ephemeral of cases.

New comment by rpdillon in "What it feels like to work with Mythos"

rpdillon — Thu, 11 Jun 2026 02:38:38 +0000

There's some truth here. But in carefully orchestrated scenarios (the minority, to be sure), it can work surprisingly well, I think.

New comment by rpdillon in "German ruling declares Google liable for false answers in AI Overviews"

rpdillon — Wed, 10 Jun 2026 10:28:25 +0000

Except the article doesn't say that, only the comment here does.

New comment by rpdillon in "What it feels like to work with Mythos"

rpdillon — Tue, 09 Jun 2026 22:31:10 +0000

This is really insightful, but I think it also extends to making the project either low stakes or low complexity. I have this lurking feeling that the preferable architecture for software will change as a result of LLMs because they're good at working on low complexity modular components more than they are on high complexity million-line code bases.

New comment by rpdillon in "Apple decided not to roll out Siri in EU after denied request for exemption"

rpdillon — Tue, 09 Jun 2026 21:26:30 +0000

It's simply a prioritizing time to market over a global release. You tend to release into the most restrictive environments last and the most forgiving environments first, for obvious reasons.

New comment by rpdillon in "Apple decided not to roll out Siri in EU after denied request for exemption"

rpdillon — Tue, 09 Jun 2026 21:19:27 +0000

Its because the standard product development strategy is to get the product into the hands of users to determine value and iterate based on feedback.

The EU has rules that are expensive to implement correctly, so if you want early feedback from users, you release elsewhere first. It's a very rational way to approach it.

New comment by rpdillon in "Siri AI"

rpdillon — Tue, 09 Jun 2026 14:07:25 +0000

I think even if what you say is true, it doesn't address parents' point that both humans and machines regurgitate what they've consumed.

But I'd also want to point out that the way you're characterizing an LLM planning a trip doesn't have any structure to it, which indicates that in your scenario you're not using any kind of harness. I've been amazed at how capable even 30 billion parameter models are when I put them inside of a harness that provides structure and task management. If you consider that scenario, especially with the ability to search the web and use skills, suddenly the LLM looks a lot more like what the human process looks like.

New comment by rpdillon in "Ask HN: What are tools you have made for yourself since the advent of AI?"

rpdillon — Mon, 08 Jun 2026 23:17:21 +0000

These are really cool - I've been meaning to return to my pikchr-in-wasm IDE experiment. I only ever built a PoC, and that was before AI coding. I should pick it back up and polish it a bit, but after seeing your DiagramIDE, I might just use that.

This whole thread is turning HN into my favorite app store. Good stuff.

New comment by rpdillon in "Ask HN: What are tools you have made for yourself since the advent of AI?"

rpdillon — Mon, 08 Jun 2026 23:14:08 +0000

How about critiquing? Like, the moral equivalent of an editor pointing out issues and/or suggesting alterations? I think it would still be the student doing all the things you pointed out, but I suspect there's a fair amount of leeway in the interpretation of each of those.

New comment by rpdillon in "Anti-social: It's fads, not friends, which now dominate social media feeds"

rpdillon — Mon, 08 Jun 2026 15:27:00 +0000

Sure, but deconstructing the platform to look at the engagement points is also useful. Some things that I think set HN apart in a good way:

The lack of any kind of personalization whatsoever on Hacker News is a huge differentiator. There are no notifications, so if you want to find out if somebody replied to you, you've got to go check. Everybody's front page is exactly the same. There are no direct messages. There are no in-line images or videos or even emoji. The feed is not endless. There is no targeted advertising. There are no reactions to posts other than upvote/downvote.

I guess you can lump HN in with Instagram and TikTok, but it just feels like a very different product, in ways that are relevant to the analysis of whether its existence is a net positive for society.

New comment by rpdillon in "The gamers taking on the industry to stop it switching off games"

rpdillon — Sun, 07 Jun 2026 20:21:24 +0000

This is the reason I buy many games twice: once when they release on Steam, and again (if I loved them) when the come out on GOG, with Doom Eternal being the most recent example, but also Doom 2016, Skyrim, Dying Light, Tomb Raider, etc. I want to reward the devs with money for giving me a copy that doesn't need servers to run. It's work to produce it, and it has value for me, so I don't mind paying. CDPR is particularly good about this (obviously, since they run GOG), but the fact that such an amazing AAA company has this stance is amazing to me.

New comment by rpdillon in "Zeroserve: A zero-config web server you can script with eBPF"

rpdillon — Sat, 06 Jun 2026 21:10:26 +0000

These projects are fascinating, and I referenced them in a nearby comment about static hosting from archives. I need to try the latest versions to see how they work at higher scale (more data in the archive).