jc4p

Do you actually find that for some use cases foundation models work better than AutoARIMA or etc? I do a lot of ecom projections and it’s never been the case for me, most of my cases are best with seasonal naive.

Would love to know actual time series foundation model use cases!

I've been spending a lot of time with my nephews working on Scratch and Snap.

They also attend local classes and their teachers switched from Snap to this which is way simpler and stickier with my nephews: https://www.microsoft.com/en-us/makecode

Just saw your edit -- I'm afraid to open source the code before refactoring it but if you reach out at [email protected] I'll send you the full ZIP!

Sorry I don't understand, you're saying the direct providers aren't the canonical source you'd recommend?

If I was running these on my own machine or GPU wouldn't the argument then be "Well you didn't use the real providers?"

For the record I started doing this approach because the Kimi team released this which was shocking to me: https://github.com/MoonshotAI/K2-Vendor-Verifier

Thank you so much for this detailed answer!! Excited to dig into this world more :)

I was using the same harness for each run, the difference is from when I was running the harness locally on my machine before I pushed up the full runs.

Great point!

When I found the original exploit in an app I researched it took me around 15 minutes and some assistance from Claude.

For this project I gave myself the weekend + parts of Monday, so around 20 hours of dev time — at my standard rate that’s ~$5,000 of dev time.

I agree fully and hope someone else is able to do this test! For me it was a matter of cost and quotas that stopped me from changing to a new account.

Also just to mention:

Claude guardrails —> that session terminated.

GPT guardrails -> your whole account is slowed down.

Thank you for your note! As I mention in the post this is not scientific at all.

I'm very curious how you would do multiple runs of multiple models in a "work alongside the model" manner?

Thank you! I’ve been fine tuning tiny Llama and Gemma models using transformers then exporting from the safetensors that spits out — My main use case is LLMs but I’ve also tried getting YOLO finetuned and other PyTorch models running and ran into similar problems, just seemed very confusing to figure out how to properly use the phone for this.

I do a lot of AI work and right now the story for doing LLMs on iOS is very painful (but doing Whisper or etc is pretty nice) so this is existing and the API looks Swift native and great, I can't wait to use it!

Question/feature request: Is it possible to bring my own CoreML models over and use them? I honestly end up bundling llama.cpp and doing gguf right now because I can't figure out the setup for using CoreML models, would love for all of that to be abstracted away for me :)

They're anonymous usernames the app had them make and they were told don't use anything shared elsewhere and I googled and there's not any uniquely identifiable people from any of them.

They seem generic enough that I think it's okay, but you're right there is no need in including them and I should've caught that in the AI output, thank you!!

Yes! haha! But hopefully I have a good enough support group and connections that I'll be ok if that happens, I just really wanted to prove that they were not being honest when they said it was data prior to 2024.

Hi all, i'm the security researcher mentioned in the article -- just to be clear:

1. The leak Friday was from firebase's file storage service

2. This one is about their firebase database service also being open (up until Saturday morning)

The tl;dr is:

1. App signed up using Firebase Auth

2. App traded Firebase Auth token to API for API token

3. API talked to Firebase DB

The issue is you could just take the Firebase Auth key, talk to Firebase directly, and they had the read/write/update/delete permissions open to all users so it opened up an IDOR exploit.

I pulled the data Friday night to have evidence to prove the information wasn't old like the previous leak and immediately reached out to 404media.

Here is a gist of Gemini 2.5 Pro summarizing 10k random posts: https://gist.github.com/jc4p/7c8ce9a7392f2cbc227f9c6a4096111...

And to be 100% clear, the data in this second "leak" is a 300MB JSON file that (hopefully) only exists on my computer, but I did see evidence that other people were communicating with the Firebase database directly.

If anyone is interested in the how: I signed up against Firebase Auth using a dummy email and password, retrieved an idToken, sent it into the script generated by this Claude convo: https://claude.ai/share/2c53838d-4d11-466b-8617-eae1a1e84f56

And here's the output of that script (any db that has <100 rows is something another "hacker" wrote to and deleted from): https://gist.github.com/jc4p/bc35138a120715b92a1925f54a9d8bb...

thank you so much!!! i should’ve put that final sentence in my post!

i've been trying to keep up with this field (image generation) so here's quick notes I took:

Claude's Summary: "Normalizing flows aren't dead, they just needed modern techniques"

My Summary: "Transformers aren't just for text"

1. SOTA model for likelihood on ImageNet 64×64, first ever sub 3.2 (Bits Per Dimension) prev was 2.99 by a hybrid diffusion model

2. Autoregressive (transformers) approach, right now diffusion is the most popular in this space (it's much faster but a diff approach)

tl;dr of autoregressive vs diffusion (there's also other approaches)

Autoregression: step based, generate a little then more then more

Diffusion: generate a lot of noise then try to clean it up

The diffusion approach that is the baseline for sota is Flow Matching from Meta: https://arxiv.org/abs/2210.02747 -- lots of fun reading material if you throw both of these into an LLM and ask it to summarize the approaches!

Amazing idea, do you think this should be a freeform text field the user can enter to add their own prompts to or should it be a checkbox/select on the homepage so the user can pick from a limited set?

I just added Romanian for you -- here's the entire diff for adding a new language (as long as it's in OpenAI's training data) -- https://images.kasra.codes/romanian_diff.png

Please let me know if it works, and I'll definitely work on adding in instructions for the expected interactivity, thank you!

Thank you so much!! While the transcription is technically in the API it's not a native part of the model and runs through Whisper separately, in my testing with it I often end up with a transcription that's a different language than what the user is speaking and the current API has no way to force a language on the internal Whisper call.

If the language is correct, a lot of the times the exact text isn't 100% accurate, if that's 100% accurate, it comes in slower than the audio output and not in real time. All in all not what I would consider feature ready to release in my app.

What I've been thinking about is switching to a full audio in --> transcribe --> send to LLM --> TTS pipeline, in which case I would be able to show the exact input to the model, but that's way more work than just one single OpenAI API call.

Thanks so much for trying it out! The realtime API is actually very cheap especially for short connections, for each user who uses it 30 minutes a day every day in a month it costs me ~$5 and I assume the average user is going to use it way less than that (although i have 0 users right now haha)

Hi! I have a WIP of this over at https://talktrainer.app/ -- I just added Dutch to it.

It uses OpenAI's realtime API to simulate either a tutoring session (the speaker will revert to English to help you) or a first date or business meeting (the speaker will always speak the target language)

You can see the AI's transcriptions but not your own, limitation of the current OpenAI API but definitely something I can fix.

The prompts are like this: https://gist.github.com/jc4p/d8b9d121425ec191d62602d8720eeed... and the rest of it is a Nextjs app wrapped around the WebRTC connection.

I'm not fully in love with the app so I'd love any feedback or hearing if it works well for you -- It doesn't have a lot of features yet (including saving context) and if you bump into the time limit just open it up in incognito to keep going.

This is super cool (i do a LOT of react native after years of doing native development) -- one thing I'd love: Adding easy upgrades, right now if I leave a RN app for a while getting it to compile on latest iOS/Android again requires a lot of manual labor and reviewing rn-diff-purge. Good luck on the project!

Autodesk bought EAGLE in 2016, so just about ten years between acquisition and discontinuation.

EAGLE is the first PCB design app I learned (and had a harder onramp than React) so this is sad, but it is important to note that most hobbyists have already switched over to KiCad: https://www.kicad.org/

love running into old friends on HN, so some very stale info from someone who hasn’t worked at SO in many years: while the job board was differentiated and nice from the programmers side, it’s really difficult to convince recruiters to use a new system with new rules. most (not all ofc) just want to spray and pray.

as a job seeker being told “you’re gonna have the upper hand here” is amazing, as a recruiter it makes it very difficult to sell to unless you really own the market.

Why would they offer him a job right before saying that? Some of it could be true, absolutely, but it seems more emotionally manipulative than anything else, to me. The CEO was mad so he said something.

This wikipedia list is missing a lot of APIs, it seems more related to products -- For example, the QPX API (which I used) was shut down this April and is nowhere in that list. Your numbers are lower than they should be.

Sorry I was mostly going off the google results. Your library looks really nice, the SCSS variable changing is awesome!

The framework being advertised as a minimal lightweight framework and having its size compared to the biggest frameworks is a bit off-putting to me.

I get that they wouldn't want to advertise their competitors, but the comparison matrix on the page makes mini.css seem tiny, where as a google search shows it's the biggest popular framework like this.

Mini.css is 7KB gzipped, Milligram[0] (the first google result I see for "minimalist css framework", mini.css is second) is 2KB gzipped. Pure.css[1] (the third result) is 3.8KB gzipped.

[0] - https://milligram.github.io/ [1] - https://purecss.io/