HN user
mail@cadocary.com
I’ve built a small suite of honey-do automations for public defense attorneys in Washington (selenium scripts and OS calls to do docket intake / organization)
Currently packaging as a windows desktop app, recording a demo video, and then distributing
Very cool. I jumped in here thinking it was gonna be something else though: a packaged service for distributing on-prem model running across multiple GPUs.
I'm basically imagining a vast.ai type deployment of an on-prem GPT; assuming that most infra is consumer GPUs on consumer devices, the idea of running the "company cluster" as combined compute of the company's machines
I'm impressed, how did you determine it was a botcomment?
Commenting because I can't get everyone's email; if you're reading this thread and would be interested in an automation engineering conference please email me at mailto:mail@cadocary.com
Yes; post was updated after my comment
In my experience chatGPT solves the challenge handily, but it does get to the level of problem that GPT requires my code review for, so either I'm not as good at using chat as I could be or they've found the point at which someone has to prove they can actually do code review to filter; nice.
I don't see how I (or LinkedIn) is making you do anything? LinkedIn is a place I can post data. I choose to do so in an attempt to market my resume. I fully expect that the data I post on LinkedIn's server becomes and is the property of LinkedIn, and wish it was more effective at extracting value from it?
Because LinkedIn is less effective than I'd like, I support 3rd parties scraping the data I posted there, again on the hope that they'd be more successful at marketing that data, which I would benefit from as the data is my resume.
Okay, so 30 million people, or more conservatively 16 million people. Same question, and before you say anything, don't be condescending.
ETA: Maybe most conservatively, let's use only the % uniquely included in U-6 and excluded in standard, or 14.4 million people. I'll claim these 14m people are the "gatekept from full employment" in that they don't qualify as narrowly unemployed unless you include "all people marginally attached to the labor force, plus total employed part time for economic reasons"(1)
Same question. 14m people who are being excluded from labor, are they not free to attempt to generate income via means other than labor, lest they suffer the judgement of Abraham?
In other words, is Abraham hiring? If not, are the people he refuses to employ meant to accept serfdom to preserve their soul?
OK, so what is to be done when access to labor is gatekept? If you can't have a job, or can't benefit from a job beyond "give a man a fish, he eats for a day", are you not meant to look for an opportunity to generate income outside of working?
I sure do! If LinkedIn can't market my resume to open roles then letting recruiters roll their own scrapers against it is the next best thing. I understand that LI owns my data, I just wish they were effective in using it!
(edit: "my" data, as in the data I post there.)
Location: Washington State, USA
Remote: on-site, hybrid, remote
Willing to relocate: Yes
Technologies:
- Frameworks & Libraries: Selenium, BeautifulSoup, Flask, Express.js, React, SemanticUI, MUI
- AI & Data Tools: PyTorch, llama.cpp, LoRA fine-tuning, Pandas, NumPy, Matplotlib
- Cloud & DevOps: AWS (EC2, RDS, SES, S3, Route 53), Docker, Kubernetes, Git, GitHub Actions, Tailscale, NixOS (yes, flake.nix is a DevOps tool.)
- Databases: PostgreSQL, MySQL, SQLite, MongoDB, DuckDB, RocksDB
- Automation & Scripting: RPA tools, PDF form automation, Web scraping, CLI utilities
- Systems & Platforms: NixOS, Linux, Windows, macOS, VirtualBox, FFmpeg
- Networking & Protocols: TCP/IP, Wi-Fi, Bluetooth/BLE, REST APIs, WebSockets
- Other Tools: PyInstaller, NSIS, Chrome WebDriver, PLC, FANUC, HMI, Oscilloscope, Function Generator
Résumé/CV: douglasmckinley.com
Email: mail@cadocary.com401 {"ok":false,"error":"X-Application-Key header is required and must be valid"}
Can't find a way to get a key on your website or in your bio?
A webscraping / data pipeline to get the .pdf "Explanation of Benefits", "Proof of Coverage", and "Drug Formulary" for every Medicare Advantage plan in the USA
These docs are gonna be used in a product for medicare brokers (if you are/know one please reach out open enrollment starts Nov.1!), and the pipeline is horizontally scalable to ingest updated 2026 plans overnight @ start of open enrollment (though some companies are posting updated plans earlier)
There are some clever tricks at play but mostly it's bog-standard browser automation; I'm also in an interview process with 2 entities (one funded startup and one massive corporation) talking about web automation roles, and while it's frustrating that they're moving so slowly it's working out to give me time to build this well.
I see your point for technical users, what I'm imagining is something where I install NixOS, and then download LibreOffice via a GUI, as is done on Windows/Mac/Most Linux tbh
I don't think an LLM is required here; for a very significant subset of nixpkgs I think updating the configuration.nix is a deterministic process.
Even assuming this is true, I do not, and will never, care.
The machine is a collection of well organized rocks over which electrons pass. I am a human.
Google is a global corporation that turns cyberstalking into money. I am a human.
Lolz are the highest aspiration available to computers. God forbid I ever lost sight of that.
I’m reading your summary versus the other article here, but it seems like for writing code, Claude would be the clear winner?
When chat breaks apart for me, it’s almost always because the context window has been overflown and it is no longer remembering some important feature implemented earlier in the chat; it seems based on your description that Claude is optimizing to not do that.
Thanks for this; I'm still not getting it, I think because I haven't really deployed software at scale before? It's a CI/CD solution?
I can see how this would be instrumental for coordinating hundreds of developers across tens of products, but as someone who's only ever worked as tens of developers on one product it seems like a lot of overhead for really no ROI; what's the difference between using this product and having a DevOps team that coordinates CI/CD in-house? Is Palantir's edge here being the out-of-house devops team for any large enterprise?
You have reached goal state of what I'm getting at here:
https://news.ycombinator.com/item?id=45160603
Do you think a "batteries-included" NixOS install could be developed for users like me? I feel like a NixOS with libreoffice, zoom, etc. could be a drop-in windows11 replacement for a lot of users.
Maybe the key feature would be a graphical installer; something that presents an app-store iterative-install UI and in the background handles configuration.nix, so nontechnical users don't have to use the terminal or wrap their head around declarative config / version control to get a reproducible desktop that they can simply install on new machines.
Reading this comment I can't help but imagine a high school student using the same pattern to respond to an "open period" being changed to "study hall" with mandatory in-library presence; which is not to dig on you, just to raise the idea that maybe k-12 education really is a conspiracy to train people to sit in factories.
How/can we "montessori-fy"?
Sick.
The first feature I wanted to add was a "Quote from xometry" panel; I cloned/GPT'd for a couple minutes and found that actually adding this would mean pulling tricks with selenium that don't scale.
Have you reached out to xometry/hubs.com/other Print-aaS companies for a partnership?
"set up a cloud instance ahead of time and just boot it up"
That sentence is the closest to what I'm imagining--the thin client to personal computer, your use case is interesting (hackathon peripherals); I'm thinking of this as a lifetime-durability play.
Basically what if my home's NAS could also serve all my applications, and at any given time I really just have a KVM into it. If we assume network is fast enough, that would achieve the UX I want, which is that the KVM is effectively disposable and the personal computer stuff all stays intact.
That’s part of but not fully it, thanks for the validation that it’s not just me re: onedrive.
I think partially I’m wrestling with the almost philosophical question of whether Linux is viable for consumers, because ideally I’d be a consumer, but I’m so frustrated by switching costs between windows/mac that surely open source has to be the solution, and loop from this point.
There are, in my mind, 3 kinds of data loss when my laptop breaks: 1) passwords 2) files 3) applications
(1) and (2) are at least in theory solved by simply embracing vendor lock in. In practice you really can just keep buying Apple forever, or figure out OneDrive for real and it will work.
(3) is where I think I’m breaking with the current state of the industry, but it is not mandatory imo that I wouldn’t be able to get my vscode, pgadmin, cad application, office suite, etc. all downloaded with their config intact on a new machine.
Location: Seattle, WA
Remote: Yes, prefer in-office though
Willing to relocate: No
Technologies: Fullstack react, express, node, postgres; data pipelines python, kafka; testing selenium
Résumé/CV: http://www.linkedin.com/in/douglasemckinley/
Email: douglas.e.mckinley@gmail.com
Don’t you already want to hit the baseball as hard as you can at every opportunity? Just with the caveat that you need to develop skill with using a one-size-fits-all bat? Is bunting and going for balls that big of a deal currently, that a player who could rock one into left field would decide not to?
Surely any strategy around loading up bases to stack the deck for your strongest hitter remains, it seems like this levels between hitters more than from hitter to pitcher?
Very encouraging, thank you
Are you hiring? I recently did a very similar coding challenge for vencover.com, and if there’s an opportunity I’d love to contribute
My struggle with all these tools is that it’s been universally easier to type “summarize this video <link>” into chatGPT than to even keep track of the link to service. I desperately want a openAI-powered business to be viable, but it’s a snake eating its tail
My worldview was challenged by linked Wikipedia page re: The American Dream, specifically this quote:
“ a phrase referring to a purported national ethos of the United States: that every person has the freedom and opportunity to succeed and attain a better life.[2] … different meanings over time. Originally, the emphasis was on democracy, liberty and equality, but more recently has been on achieving material wealth and upward mobility.[4]”
It seems like the article—and many of popular complaints (including mine)—are primarily grappling with the lack of upward mobility. Versus the older meaning, “democracy, liberty and equality”, which IMO would be a much more severe loss.
An argument could be made that loss of material equality threatens access to democracy and liberty, but imo the article—and charity recommendations-are focused on bandaids, where the [Citizens United](https://en.wikipedia.org/wiki/Citizens_United_v._FEC) decision is the known mechanism by which inequality threatens democracy and liberty.
When/if I think of a corporation as a legitimate group of citizens, (much less dissonance when I think of a labor union this way), I wonder if the democracy and liberty are better protected by allowing inequality to be weaponized against the government. The “fact” that weaponizing inequality against the government is the effect of Citizen’s United remains unchanged?
Can anyone recommend a CUDA equivalent of (2)? That’s a spectacular learning resource and I’d like to use a similar one to upskill for CUDA
Lmao yeah maybe I’m showing my age; stop trying to make Fetch happen again
Being an employer (small medical business, 10ppl)
This may not be an edge case in terms of # of employers that fit your description, but is very likely an edge case in terms of # of employees with an employer that fits your description.
You’re the exception that proves the rule.
—————
Edit: Looking into it more, the margin is closer than I assumed; ~half of employees work for a “small business”, that term meaning <500 employees (1). Of those, ~80% are for a “small business” with <10 employees (2).
1: https://advocacy.sba.gov/2023/03/07/frequently-asked-questio....
2: https://www.pewresearch.org/short-reads/2024/04/22/a-look-at....
So you’re representing the majority of half the employment marketplace. That said I’d still argue that conversations about labor relations are focused on the subset of companies with many hundreds, thousands, and/or hundreds of thousands of employees.
—————
Interestingly, the majority of small business revenue is generated by the extreme minority of small businesses—per (2), the “small businesses” with >50 employees (so, 50<employees<500) represent just 3.3% of “small businesses”, but generate 53% of the revenue among all “small businesses”