Not disagreeing per se, but a quick look at the installation instructions confirms what I assumed:
Yeah, you can run a highly quantized version on your 2020 Nvidia GPU. But:
- When inferencing, it occupies your "whole machine.". At least you have a modern interactive heating feature in your flat.
- You need to follow eleven-thousand nerdy steps to get it running; my mum is really looking forward to that.
- Not to mention the pain you went through installing Nvidia drivers; nothing my mum will ever manage in the near future.
... and all this to get something that merely competes with Haiku.
Don't get me wrong - I am exaggerating, I know. It's important to have competition and the opportunity to run "AI" on your own metal. But this reminds me of the early days of smartphones and my old XDA Neo. Sure, it was damn smart, and I remember all those jealous faces because of my "device from the future." But oh boy, it was also a PITA maintaining it.
Here we are now. Running AI locally is a sneak peek into the future. But as long as you need a CS degree and hardware worth a small car to achieve reasonable results, it's far from mainstream. Therefore, "consumer-grade hardware" sounds like a euphemism here.
I like how we nerds are living in our buble celebrating this stuff while 99% of mankind still doomscroll through facebook and laughing at (now AI generated) brain rot.
I've actually been using the Gemini app more because it auto-deletes old history. I like using LLMs without thinking this is going to stick around forever.
Models are relatively interchangeable for day-to-day use anyway.
Can you explain how you achieved it? I don't find it in your repo, I am too tired probably? However, I tried something different but stuck eventually with the receiving part on Android. One cannot just pull files from somewhere and put them into the Obsidian folder. I mean, not without efforts.
Thanks for being my very first HN reply — genuinely appreciate it.
The short answer: I cheated by staying in the Apple ecosystem.
The sync has two layers:
- *macOS ↔ iOS*: iCloud handles it natively. Obsidian's iCloud vault is just a folder, both devices read/write the same files, syncs in seconds.
- *macOS ↔ GitHub*: A launchd job runs `git add + commit + pull --rebase + push` every 5 minutes. That's the entire mechanism — it's in `scripts/sync.sh`.
So there's no magic — iCloud does the heavy lifting for mobile, Git handles versioning and cross-platform backup.
Your Android problem is real and hard. Android doesn't give apps transparent filesystem access the way iCloud vaults work on iOS. I haven't solved that yet. The closest path I can think of: Termux + git clone + symlink into Obsidian's storage, but that's far from "zero effort."
This is worth thinking about more deeply. If I find a clean solution, I'll come back and share it here.
Problem is, that just having a Claude subscription doesn't make you productive. Most of those talks happen in a "tech'ish" environments. Not every business is about coding.
Real life example: A client came to me asking how to compare orders against order confirmation from the vendor. They come as PDF files. Which made me wonder: Wait, you don't have any kind of API or at least structured data that the vendor gives you?
Nope.
And here you are. I am not talking about a niche business. I assume that's a broader problem. Tech can probably automate everything and this since 30 years. Still business lack of "proper" IT processes, because at the end every company is unique and requires particular measures to be "fully" onboarded to IT based improvements like that.
Thing is, unless the order confirmations number high in the thousands, most businesses could just hire an old lady with secretary experience to crunch through that work manually for a fraction of what it would cost them to get the whole system automated.
I've seen this play out with time sheets. Every day factory floor workers write up every job they do on time sheets, paper and pencil, but management wants excel spreadsheets with pretty plots. Solution? One old typist who can type up all the time sheets every day. No OCR trouble, no expensive developer time, if she encounters illegible numbers she just cross references against the job fliers to figure it out instead of throwing errors the first time a 0 looks like a 6.
first I was like "What but why? You don't save any space or what's that excercise about" then I read it again and it blew my mind. I thought I knew everything about ASCII. What a fool I am, Sokrates was right. Always.
That’s essentially what I’ve experienced, call it 'anecdotal evidence.'
I had a long, ongoing, and very upsetting interaction with a bigger German company. Since I have experience in data privacy and GDPR, I eventually started thoroughly crawling their entire online presence for infringements.
I found a significant number of issues and compiled a very extensive report. At first, they were completely dismissive. It was only after I issued formal legal warnings that an actual lawyer contacted me and promised to fix the issues.
Most of the GDPR violations were simply sloppy, though some were genuinely ignorant. It’s wild that we are eight years past 'Year Zero,' and while everyone is constantly talking about data privacy, these gaps still exist.
Some of them eventually has been fixed after my report, silently of course. phhh...
Mom and pop businesses with limited IT skills are not collecting emails and private information. At worst they’d be using some external service (e.g. Mailchimp) which does it for them, and those have an obligation to be familiar with the law.
The GDPR really isn't that hard to follow, for a "mom & pop" business, it really comes down to:
* Limit data retention — Don't keep personal data longer than necessary
* Honor data subject rights — Allow individuals to access, correct, delete, or port their personal data
Simply, don't collect personal information if you don't need it. If you do need it, add a delete button.
That's not a good link for a list of past threads since the idea for the latter is to include only the ones with interesting comments.
However, it looks like a good article that could use a repost! Just not soon, since we want to give enough time for the hivemind caches to clear :) - if you want to repost it in (say) a month or two, email us at [email protected] and we'll put it in the SCP (https://news.ycombinator.com/item?id=26998308).
Not disagreeing per se, but a quick look at the installation instructions confirms what I assumed:
Yeah, you can run a highly quantized version on your 2020 Nvidia GPU. But:
- When inferencing, it occupies your "whole machine.". At least you have a modern interactive heating feature in your flat.
- You need to follow eleven-thousand nerdy steps to get it running; my mum is really looking forward to that.
- Not to mention the pain you went through installing Nvidia drivers; nothing my mum will ever manage in the near future.
... and all this to get something that merely competes with Haiku.
Don't get me wrong - I am exaggerating, I know. It's important to have competition and the opportunity to run "AI" on your own metal. But this reminds me of the early days of smartphones and my old XDA Neo. Sure, it was damn smart, and I remember all those jealous faces because of my "device from the future." But oh boy, it was also a PITA maintaining it.
Here we are now. Running AI locally is a sneak peek into the future. But as long as you need a CS degree and hardware worth a small car to achieve reasonable results, it's far from mainstream. Therefore, "consumer-grade hardware" sounds like a euphemism here.
I like how we nerds are living in our buble celebrating this stuff while 99% of mankind still doomscroll through facebook and laughing at (now AI generated) brain rot.
(No offense (ʘ‿ʘ)╯)