Hacker Newsnew | past | comments | ask | show | jobs | submit | carterschonwald's commentslogin

theres a certain amount of financial security needed for folks to do this kinda thing.

Of course. But software developers are, on the whole, pretty damn well compensated. As a group we have a lot of freedom to make these sorts of financial decisions.

Yes, some members of the group don't have that freedom. But that doesn't excuse the ones who do.


Even if everyone had enough financial security (or realized they did), you need coordination, people are never going to independently and spontaneously turn down these jobs at evilcorp en masse - you need coordination, a.k.a strong unions which can actually effect the policies of the employer.

The article addresses that:

> If the answer is the salary, say it louder, and do not force yourself to make excuses.


No. This is one of the many excuses that people employ to keep feeding the beast.

You need to have money to do the right thing? Odd take. Plenty of people with much less have sacrificed much more throughout history.

When you make a sacrifice, it has to make sense, and must have some chance for success. In this case it needs to be recognized by others, and a contemporary person can hardly understand any other sacrifice than monetary.

"Now that my RSUs have been exercised, here is the problem with _________"

the current white house really fucked over all the smaller domestic manufacturers. and scientists. and um america basically

Everything.

If you are a builder, there is no sane way to bid on contracts with such hugely variable COGS.

And how the HELL do you even attempt to fix the US housing supply if you are tariffing Canadian lumber? You can't.


Guess this is what happens when you elect a group that prioritizes enriching themselves over ensuring you don't starve in the street like a filthy animal.

Likewise, how do you attempt to fix US manufacturing if you are tariffing Canadian aluminum?

- US imports 20% of its lumber from Canada vs. 70% local lumber.

- Raw lumber costs are about 1% of the price of a house in California.

- Lumber prices were much much higher during the pandemic and are actually not that high right now: https://tradingeconomics.com/commodity/lumber (it's at 9 months low and cheaper than in 2018).


It's not just lumber.

There is a 24-45% tariff on building supplies, steel, aluminum, and cement.

> Raw lumber costs are about 1% of the price of a house in California.

California is fucking weird then, because it's 16% or more other places.

https://www.txrac.com/blog/cost-to-build-a-house-from-scratc...


As for the housing... stop using lumber maybe? Us Europeans build houses out of bricks and mortar or outright concrete.

Lumber framing only makes sense if you want something dirt cheap (and yet, US matchsticks and cardboard houses go for millions of dollars?!) or something that needs to be earthquake resilient and affordable.


I have a video for this.

It's because of our earthquakes and the severity of our storms. The lumber frame and design of construction flexes and sways when stone and brick would break (tornados are different story). Likewise when they do break, and they absolutely will given some of these forces, it doesn't bankrupt people and companies when trying to recover from those events. Bigger buildings of concrete, stone and brick here are now built very different, sometimes on giant metal plates with actual springs.

Here you go:

https://youtube.com/shorts/jFph1d2WR3s

Btw:

> or something that needs to be earthquake resilient and affordable

https://earthquake.alaska.edu/earthquakes


What is the deal with europeans who have zero knowledge of construction having a weird grudge against wood framing?

> europeans who have zero knowledge of construction

I actually worked in construction and now studying electrical engineering, focus residential.

> weird grudge against wood framing

Well, it's cheap to build but sold to your population at prices that have us ROFLing, it needs to be treated with all kinds of nastyness to survive stuff like termites, a lot of construction is incredibly shoddy (watch some Cy Porter), it's not durable like our houses that last many decades to centuries... choose your pick.


I've lived most of my adult life in wooden structures that are over 100 years old. They will stand for another 100 as long as waterproofing is maintained (and a masonry building will be severely damaged as well if it is not waterproof). They have been through multiple earthquakes and are easy to repair if damaged, by opening up one of the "cardboard" walls and nailing in some more wood pieces. When they are built in wildfire-prone localities, they are coated in stucco and given concrete roof shingles, making them as fire resistant as masonry buildings. Houses are expensive in the US because the land is expensive here (location location location), and because there is shortage of skilled labor due to the high employment rate. Areas of the US with levels of economic performance similar to Europe have similarly cheap houses.

> it's not durable like our houses that last many decades to centuries...

And none of your homes would last more than 50y here before an earthquake, tornado, or hurricane destroyed them.


Ignorance, mostly. They don't understand how the layers and designs work together because they don't live here.

i actually have a harness setup that prevents role confusion from happening in a much more robust and interesting way.

hoping to launch a nice commercial version as a saas with some compelling unique features in the next month or teo


looking for any kind of testing or help? it’s been probably 10+ years since we were in the same irc channels (cmr/KB1PKL). email is in profile :)

I'm assuming that vagueness is intentional, but can you tell us the general approach? Is it built on top of inference APIs as a primitive, or is it a novel approach like I'm talking about that inspects the model's internals?

i do have some ideas that could be done at the model arch/inference level, but a shocking amount of stuff works even just as clever formatting

good. theyll actually be more reliable if they dont have as much brain damage.

precisely. we jam their few-dozen-slot global workspace with incoherent posttraining.

i think the lower bound on the end state is there cant be opaque reasonibg steps ever.

It is distinctly likely that visibility and general reasoning at humanlike speed and efficiency is impossible. That is reasoning at the token level and at the meta level don't have a one to one representation that can be interpreted while using the same amount or less energy.

this is just not good science or engineering. ok as a student project i guess.

im pretty literate in intellectual property, but im pretty confused about the no looking at binary code artifacts bit.

TL;DR:

1. We want to guarantee our work is not a "derivative work" of anything Apple wrote.

2. If we look at any Apple binaries, there's no way prove that our code didn't borrow from Apple.

3. Since we didn't look at any Apple binaries, then there's no way our work can be a derivative work-- we didn't even look at their stuff.


Wouldn't using OpenAI break the cleanroom?

That is the debate. And it is hotly contested.

The legal framework around all this is way too immature to make a definitive statement one way or the other. Anyone who says otherwise is speculating.

The only way to clarity will be legislation or court cases, because it really could go either way.


Not hotly contested by impartial legal types. There is a -6543.21% chance the Roberts’ Court does anything to hold ai back.

If using the OpenAI models to develop this taints you with Apple insider knowledge, then so is developing any macOS, iOS, Android or Windows apps around that are using any LLM.

You can't claim that some work is necessarily tainted by existing, you also have to prove which knowledge was retrieved, used and that there was no other reasonable way to know it. But unless it is actually tried in a court in all major jurisdictions, it's just speculation from all parties (including line).

One person could also claim that someone someday might leak information to them, and even though they'll ignore it, since they'll be exposed, they are tainted and can't contribute anymore. It'd be absurd.

A project is free to chose who can contribute, but they can't necessarily gatekeep others from doing their own project with different rules.


Yeah I see developers constantly saying they RE stuff using LLM in a clean room way and it’s definitely an opinion, not a fact. Just look at the most recent post here: https://x.com/ewninjaofficial

Did Apple share their code with OpenAI ?

Aren't they literally being sued for taking it from Apple?

Why would it? Unless you think OpenAI is training on Apple’s proprietary code?

They could claim that some employee one day used their model to ask a question about internal stuff and they trained on that.

It doesn't have to necessarily have to make sense to make a claim, but it doesn't necessarily bear any weight.


… i thought international treaties explicitly say no space weapons. ughhhh fucking wh

FTA it says the treaty covers "weapons of mass destruction". But historically (at least prior to the 2000's), there was an unspoken rule of not weaponizing space. I think that's changed in the last ~20 years and recall covered of China sending kinetic weapons into space.

And you think America cares about treaties?

Historically it did, so long as the politically privileged class of USians were among the folks that would be advantaged by those treaties.

Anti-nuke proliferation and CFC restriction treaties worked.

Treaties with native populations were ignored and violated whenever convenient; now they get occasional acknowledgement in speeches but no actual change or action.

I would’ve put space weaponry in the first category.


Laws are for the little people, obviously.

repeat after me: yaml is not a programming language. (boo, hn strips emojis, i should know that)

very disappointed that their domain specific language is just yaml. conditionals and variable binding become insane war crimes when yaml comes to town. you have a friggin llm, do better

and when we computer folks see the word language its implied that its a computing on computers language not a friggin data format.

edit: the yaml to avoid complexity that should live in the sql side can back fire, some folks I was working with last fall were evaluating if a yaml based OLAP tool would work for them, and there were some pretty gnarly gotchas from a yaml based approach. Secondarily, theres a real case to be made that having the data linkages not visible in the charting layer means that groups of related plots with different axes wont have the right data linkage without forcing a lot more ETL for what should be a quick plot if the data already fits in memory.


what would you have chosen for your DSL? Ideally something not from scratch

i'm not sure if that caveat is sensible :)

theres actually a very important reason you want it to be an actual embedded dsl or tiny programming language!

The reason why llms can code at all is the hugeeeee amount of RL based on the loop of 1 "write code", 2 get compile time or runtime errors,3 fix it and iterate. Data file formats dont have that feedback loop so models will fall off the rails faster. Writing code that fits a latent adhoc schema just wont work as well, or will require burning a lot more context.

from that perspective, it could just be an EDSL little library in the host language, or it could be a friggin little custom language with an interpreter and good error messages.


Oh to be clear

> dct validate <file.yml>

Or any of the other commands that read the file will quickly fail with any syntax or SQL issues. It’s not just an open schema.


thats important! but how do the error messages compare with a decent compiler? youre gonna have more errors in your llm corrective edits loop if you dont hew to that tier of clarifying.

to be clear: youre taking this feedback wonderfully well and i appreciate that maturity :).

im just cranky with the current opportunity market and mucking with attacking a classical markov chain conjecture while i hurry up & wait to hear back from places XOR slowly build some stuff in the harness space that i hope can hit the equivalent of sublime text tier but for harnesses (is there even a market for a prosumer/pro tier harness, im not sure!)


Fair hit on "language". I keep forgetting YAML stands for "YAML Ain't Markup Language". What we built is a declarative spec written in YAML, and that's deliberate. Think of it as HTML for boards rather than a programming language. The tradeoff is readability and structure versus flexibility.

That lack of flexibility is actually a feature for one of the bigger problems we're solving: lineage, or knowing where a number came from. If the chart layer can't transform data, the logic stays in SQL, where you can version, test and audit it.

It's also a choice about who this is for. The data community already works in SQL, Jinja and YAML every day, so there's no new language to learn.


ive been slowly working on my ideas for a proper second gen multi device harness that i want to be in the same quality regime as sublime text. unclear if theres a viable commercial market (and for a variety of reasons around user security and reliability thinking mebe saas). at the very least i want nice things for me.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: