Foreignborn

First, really cool project and love what you're doing.

I'm working on something VERY similar, but focused on a particular country, so I'm in the weeds (hah) on this and here's my feedback:

* I absolutely do not trust the data provenance here, and reading through all your links doubly so. You cannot trust claude to do analysis. Not only does it misattribute often, but fake sources and poor interpretations seep in. Without url sources tied to each claim it's inevitable that it rots. You'll have no way to keep track of claims across plants, and over time things slip.

I did see your data structs in github, but a lot of it is just LLM summaries of what it read. This isn't bad (and at some point necessary), but Claude in particularly is notorious on tasks like this, that don't pass human evaluations.

HOWEVER, that said, it's probably okay because it's not a life or death subject. Claims like "full sun" aren't very usable as is, and these can just serve as pointers for people to go research plants on their own.

----

Excited to see where you get to!

i live in shibuya and i have countless videos where it’s <40db outside.

just make the damn building with noise in mind (RC, interior insulation, gapped flooring, etc).

it’s very possible.

(btw I am as country bumpkin as they come, my village has less than 200 people so i understand rural living.)

are there tells beyond the lyrics? I swear there are a number of songs using out of “human distribution” words, trigrams, etc.

Is it overall song structure?

you’re admitting this is a dog whistle?

on behalf of the amazingly well dressed colleagues over my career, post a fit and let’s see.

I _really_ think you have an interesting tool, but the workflow loop isn't fully there.

Please let me revise or remix a suggested node. I find them extremely engaging, and I can envision ways of sort of "spinning off" even further than it's suggestions. Think Brian Eno's Oblique Strategies.

To me, this is hinting at really interesting creative processes that feel much more humane than how most LLMs work today.

these two posts (the parent and then the OP) seem equally empty?

by level of compute spend, it might look like:

- ask an LLM in the same query/thread to write code AND tests (not good)

- ask the LLM in different threads (meh)

- ask the LLM in a separate thread to critique said tests (too brittle, testing guidelines, testing implementation and not out behavior, etc). fix those. (decent)

- ask the LLM to spawn multiple agents to review the code and tests. Fix those. Spawn agents to critique again. Fix again.

- Do the same as above, but spawn agents from different families (so Claude calls Gemini and Codex).

—-

these are usually set up as /slash commands like /tests or /review so you aren’t doing this manually. since this can take some time, people might work on multiple features at once.

I value the author's optimism, but this list activates a deep fear in me. It feels like a such an American point of view to think a techno-ideal is defined simply by buying a different set of consumer electronics, many of which are just as throwaway as the ones before.

I know we can't get away from buying things--even a self-hosted homelab needs parts, and I'm not rejecting capitalism. But it feels like capitalism culture is so strong that it goes through everything people think like thread through a needle, everything they do is stitched with its color.

This makes me sad for the future my children will inherit. I want them to be excited by what comes next, the way I was excited by the N64 or the early web. But those things were exciting because they were _new frontiers_ and new stories, not because they were products.

If the only future we can envision is a curated list of retro-gadgets and subscriptions, we have lost the plot.

I love HN because everyone is so different outside of the core purpose of the site. Sometimes people reference art, or a book or something, that I'd never would think to exist.

Llainwire was my top artist listens throughout 2023, so it’s always funny to bump into reactions that feel totally different from my world/my peers.

I want to shoutout Nial Ashley (aka Llainwire) for doing this in 2023 as a solo act and doing the visuals himself as well - https://www.youtube.com/watch?v=M1ZXg5wVoUU

A shame that kid was slept on. Allegedly (according to discord) he abandoned this because so many artists reached out to have him do this style of mv, instead of wanting to collaborate on music.

it at least seems like it has a modicum of human thought, whereas this GPT drivel does not.

the cheap house thing has nothing to do with if you’re foreign and everything to do with a bank not wanting to lend money for something there’s no market for.

you can easily buy these homes with cash. i know because i have a ton of friends that have done exactly this.

if you want a cheap house in a good market like tokyo, osaka or fukoaka, then you can do so through a bank via the normal routes

I just read the book last week. What you said is not true in any useful sense. “Germans were acutely aware…” tries to reduce an entire population and years into one statement. Reality has much more color.

For the germans interviewed in the book, it seems to be true that many had read or heard about the camps or other atrocities, but (1) not the “final solution” which was not in the press and (2) there seems to be heavy desensitization from 1933-1955 when the book was written.

Aside from the tailor that had started the fire at the synagogue, the other 9 interviewees had not directly witnessed atrocities being committed, and instead focused on their personal hardships during the war.

Even though they may have been literate, the people in Mayer’s book were ignorant of the specific realities. Perhaps willfully ignorant, yes, but the nazi regime really did not give any opportunities otherwise.

—

not an expert, just reporting my notes from the book.

i highly recommend all americans read it, its not a long book. it feels eerily familiar, even though many circumstances are drastically different.

do you have a source?

when i’ve done toy demos where GPT5, sonnet 4 and gemini 2.5 pro critique/vote on various docs (eg PRDs) they did not choose their own material more often than not.

my setup wasn’t intended to benchmark though so could be wrong over enough iterations.

can you say more about world models or symbolism?

i thought world models like genie 3 would be the training mechanism, but i likely misunderstand.

These stories always have me instantly sobbing, life can be tragically unfair.

Everyone is responsible for what they deliver. No one is shipping gluttonous CLs, because no one would review them. You still have to know and defend your work.

Not sure what to tell you otherwise. The code is much more thought through, with more tests, and better docs. There’s even entire workflows for the CI portion and review.

I would look at workflows like this as augmentation than automation.

I have a team that’s somewhat junior at a big company. We pretty much have everyone “vibe plan” significantly more than vibe code.

- you need to think through the product more, really be sure it’s as clarified as it can be. Everyone has their own process, but it looks like rubber ducking, critiquing, breaking work into phases, those into tasks, etc. (jobs to be done, business requirement docs, domain driven design planning, UX writing product lexicon docs, literally any and all artifacts)

- Prioritize setting up tooling and feedback loops (code quality tools of any and every kind, are required). this includes custom rules to help enforce anything you decided during planning. Spent time on this and life will be a lot better for everyone.

- We typically making very very detailed plans, and then the agents will “IVI” it (eg automatic linting, single test, test suite, manual evaluation).

You basically set up as many and as diverse of automatic feedback signals as you can.

—-

I will plan and document for 2-4 hours, then print a bunch of small “PRDs” that are like “1 story point” small. There’s clear definitions of done.

Doing this, I can pretty much go the gym or have meetings or whatever for 1-2 hours hands off.

—-

yes, but those aren’t released and even then you’ll always need glue code.

you just need to knowingly resource what glue code is needed, and build it in a way it can scale with whatever new limits that upgraded models give you.

i can’t imagine a world where people aren’t building products that try to overcome the limitations of SOTA models

you want more options for voices that reflects all the types of people in the world. good feedback.

The next part i’m only saying because it reminds me so much of my younger self: The rest of what you said, and how you said it, has a lot of projection and insecurity.

way too many people get hung up on the idea OP led with.

what OP has is a fully self hosted, private video feed that can alert for more sophisticated events like:

- is anything happening that shouldn’t be? - are things not where they’re supposed to be? - has anything fallen? (plants, things into the pool)

good job OP. i’m going to take a look at this in the next couple months.

It’s easy to get 500-700k tokens in. I’ll drop research papers, a lot of work docs, get through a bunch of discussion, before writing a PRD-like doc of tasks to work from.

That generally seems right to me, given how much we hold in our heads when you’re discussing something with a coworker.

Try dropping the entire api docs in the context. If it’s verbose, i usually pull only a subset of pages.

Usually I’m using a minimum of 200k tokens to start with gemini 2.5.

i don’t quite understand, what makes it reverse typesetting?

my understanding is your typesetting books for responsive eink readers.

Out of curiosity, what happens to you when a less privileged person thinks the same thing about you?

San Francisco, up until covid this still happened for me too. there was a bodega like store in the mission me and other people used.

where i live now differs so that phenomenon doesn’t exist here.

please, please stop letting AI rewrite for you. i’m so tired of reading AI slop.

instead, ask it to be a disagreeable editor and have it ask questions about your draft. you’ll go so much further, and the writing won’t be nonsensical.

thanks for this, it sounds pretty good.

link for anyone else: https://canopylabs.ai/model-releases

it’s because they used an LLM to edit the prose. It has all the signature marks.

There might be good content there, but it’s full of AI slop.

i don’t have a stake in this at all, but i've seen dozens and dozens of comments of exactly what you’re asking here on HN.

People have been sharing that, plus the problems they’ve solved or products they’ve made.

Though, most of these workflows are no longer really just “chats” at like chatgpt.com. A lot of that has changed.

i’m not saying you need to do it, or change anything, but… i’ve seen people do shit with a computer lately that 6 months ago would seem weird and alien.

Tailwind has solved a lot of problems. And this is coming from someone who doesn't use it half the time.

Some thoughts about this topic, not necessarily in disagreement:

- Designers seldom think in the way you describe (explicit computer-readable rules, systems of design). A lot of designers are coming from crafts that are more implicit. It doesn't surprise me design business logic is not the topic of conversation.

- You need to update your tailwind classes to reflect your design business logic. Don't do "bg-gray", do "on-surface-emphasis" or whatever business logic. This is just a simple config change.

- You probably brought up modular scale as just one example, but modular type has become near myopic dogma of frontend engineers who haven't had any typesetting experience. A scale of 1.25 vs 1.5 is so in the weeds as to be useless. There are so many alternative ways to set type (e.g. take lessons from CJK typesetting) or innovate typography on the web (e.g. build your own leading-trim implementation!).