Hacker News: tomashubelbauer

New comment by tomashubelbauer in "Corporate America is getting hooked on open-source AI"

tomashubelbauer — Fri, 04 Sep 2026 16:28:27 +0000

I think it truly just was Opus 4.5 where LLMs became usable for coding.

New comment by tomashubelbauer in "Agent memory as a file format"

tomashubelbauer — Mon, 31 Aug 2026 17:22:31 +0000

Can an LLM be trained to understand language without remembering anything else from its training data? I thought the intrinsic knowledge and the ability to understand language were tied together.

New comment by tomashubelbauer in "MiniageOS: "Dumbphone" Version of LineageOS"

tomashubelbauer — Sat, 22 Aug 2026 19:20:35 +0000

I thought this was going to be an OS for actual dumbphones. Had me hoping I would have a reason to dig out my childhood Sony Ericsson W810.

New comment by tomashubelbauer in "Count Binface receives over a quarter of votes in Clacton by-election"

tomashubelbauer — Fri, 14 Aug 2026 20:28:57 +0000

I seriously doubt as a public servant she'd be a beacon of efficiency so I don't think she'll do more than 2k a day if this happens.

New comment by tomashubelbauer in "Mea Culpa – Dark Hours"

tomashubelbauer — Sun, 09 Aug 2026 14:41:11 +0000

Reminds me of politicians and public figures saying they got "hacked" after doing/posting something stupid and getting backlash for it. I guess "AI did it" is the new iteration on that.

New comment by tomashubelbauer in "Parsers don't have to be complicated"

tomashubelbauer — Fri, 07 Aug 2026 09:47:33 +0000

This post doesn't touch on something that makes parsers complicated no matter how simple the grammar: good error messages. Parsing a well formed input is the easy part, but not just spitting out a byte index but actually telling the user why their input is not good and what they could do to make it conform is super hard.

The Rust compiler is a common example of a compiler that does a good job here, and I think it is one of only a few.

New comment by tomashubelbauer in "Changes at Google DeepMind: Demis Hassabis from CEO to Chair, Jeff Dean departs"

tomashubelbauer — Thu, 06 Aug 2026 10:31:45 +0000

What is your position on middle of the road pain and suffering?

New comment by tomashubelbauer in "“Gravity is worth asking about”"

tomashubelbauer — Wed, 05 Aug 2026 10:19:26 +0000

I have been using macOS for I think fewer than ten years and I think the attention of detail may always have been just about the hardware... macOs and iOS have presented me with some of the clunkiest edge cases and mishandled situations and a lot of them are damn near undebuggable. I think Windows is very tasteless OS making macOS seem good in comparison but in isolation there is a lot that could a should be improved

New comment by tomashubelbauer in "Kimi K3 Now Available via Telnyx Inference API"

tomashubelbauer — Tue, 28 Jul 2026 08:48:15 +0000

FWIW Claude sometimes identifies as DeepSeek when asked in Chinese: https://x.com/stevibe/status/2026227392076018101

New comment by tomashubelbauer in "Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber"

tomashubelbauer — Tue, 21 Jul 2026 16:16:18 +0000

More like living next to an ice cream truck car park

New comment by tomashubelbauer in "Show HN: I replaced a $120k bowling center system with $1,600 in ESP32s"

tomashubelbauer — Sun, 19 Jul 2026 19:04:38 +0000

Please share more :)

New comment by tomashubelbauer in "Battery packs: Let's talk about crates, baby"

tomashubelbauer — Sat, 18 Jul 2026 13:18:48 +0000

Homebrew is also great in this regard

New comment by tomashubelbauer in "Kimi K3: Open Frontier Intelligence"

tomashubelbauer — Thu, 16 Jul 2026 20:06:18 +0000

This is a cool idea. I know I'd rather see this comment on every model release than the pelican.

New comment by tomashubelbauer in "Open source AI must win"

tomashubelbauer — Sat, 13 Jun 2026 11:00:14 +0000

I don't even need today's frontier, give me a local model I can run on my Mac comparable to Claude 4.5 as of December last year and I'll probably lose any interest in new hosted LLM advancements altogether.

New comment by tomashubelbauer in "Chipotlai Max"

tomashubelbauer — Tue, 02 Jun 2026 13:40:48 +0000

The ingredients are:

1. Bun.Cron API to run a script every minute

2. Bun.$ (Bun Shell) to execute the macOS command to take a screenshot (I do this for all connected screens at that moment)

3. Bun.Image to downscale everything to 1x in case some of the screenshots are 2x

4. Bun Shell again to run a JXA AppleScript thing to use the Vision Framework or whatever it is called to OCR the image into a file

5. Bun Shell to run the Swift compiler in the one-off eval mode with inline Swift helper that runs the Foundation Models Framework built-in LLM with a system prompt that tells it what the OCR said and instructs it to glean what may be on the screen (can't do this with JXA because the models are not exposed with ObjC APIs)

6. For each screenshot, continuously, take the previous day summary file and the last OCR/context results and produce a new summary of the day

I plan on adding extra information from the OS like the currently opened windows, currently focused window, time of day etc. into the mix, but so far it hasn't been needed. It produces reports of a good enough quality for me.

I `grep` these daily summaries whenever I need to recall a link I saw or a find what channel a message I spotted was in or take another look at that one tab I already closed, maybe re-open it by its OCR'd URL etc.

New comment by tomashubelbauer in "Chipotlai Max"

tomashubelbauer — Tue, 02 Jun 2026 13:14:47 +0000

I have a personal, fully offline and local version of Windows Recall basically, but good, made using macOS built-in OCR and LLM. The reasoning requirements are tiny (just interpret the screen based on the OCR, do rolling de-duplication and summarization), but they are non-zero. The tool is valuable to me and it being dep-free and fully offline and local just gives me a good feeling.

New comment by tomashubelbauer in "Chipotlai Max"

tomashubelbauer — Tue, 02 Jun 2026 05:40:31 +0000

If you're on macOS you can try the built in LLM which I think is similar in size. There's a project called Apfel that wraps it in a CLI. Also Chrome ships with a web API called Prompt API that gives you offline access to Gemini Nano which can do both text and images at the input. Also tiny. I've integrated these into my workflows where a tiny but non zero amount of reasoning is needed in between the otherwise fully deterministic steps.

New comment by tomashubelbauer in "Last.fm is now independent"

tomashubelbauer — Wed, 27 May 2026 16:35:28 +0000

I think Last.fm might have been a better friend finding and dating app than any of its contemporaries or anything that came after. Seems like everyone in this thread or anyone I know IRL has a story of making a good connection with someone via it. I know I'll always cherish the people I got to know on there.

New comment by tomashubelbauer in "I'm Tired of Talking to AI"

tomashubelbauer — Wed, 27 May 2026 14:47:30 +0000

I just keep repeating "operator" ad nauseam not stopping for anything other than the sound of the call transferring and that almost always works.

New comment by tomashubelbauer in "The Whole Anthropic Kerfuffle"

tomashubelbauer — Thu, 14 May 2026 13:43:04 +0000

There is basically no support to speak of. Scaling a zero is not hard. You can be paying 200 USD a month with barely any chance of ever hearing back. Your best chance of getting support from Anthropic is the same as with any other big tech company: have a Twitter following or know someone who works there.