Hacker News: antimatter15

New comment by antimatter15 in "Ask HN: Are you still using a Vision Pro?"

antimatter15 — Tue, 09 Jun 2026 18:56:25 +0000

New comment by antimatter15 in "Codex for almost everything"

antimatter15 — Thu, 16 Apr 2026 21:17:23 +0000

I remember looking trying to build something like this 6 years ago[0]. There are some interesting APIs for injecting click/keystroke events directly into Cocoa, and other APIs for reading framebuffers for apps that aren't in the foreground.

In particular there was some prior art that I found for doing it from the OpenQwaQ project, which was a GPLv2 3D virtual world project in Squeak/Smalltalk started by Alan Kay[1] back in 2011.

If I recall correctly, it worked well for native apps, but didn't work well for Chromium/Electron apps because they would use an API for grabbing the global mouse position rather than reading coordinates from events.

[0]: https://github.com/antimatter15/microtask/blob/master/cocoa/... [1]: https://github.com/OpenFora/openqwaq/blob/189d6b0da1fb136118...

New comment by antimatter15 in "MEMS Array Chip Can Project Video the Size of a Grain of Sand"

antimatter15 — Mon, 13 Apr 2026 18:16:40 +0000

This reminds me of the original patents that Magic Leap had, which involved pumping light through a single optical fiber that was wiggled by piezoelectrics into a spiral to project light (https://kguttag.com/2018/01/06/magic-leap-fiber-scanning-dis...).

New comment by antimatter15 in "Data has weight but only on SSDs"

antimatter15 — Wed, 04 Mar 2026 20:54:26 +0000

Another fun calculation is that due to special relativity, a hard drive that is spinning gains a certain amount of mass due to the rotational kinetic energy and E=mc^2.

Assuming the platter is 100g, 42mm, spinning at 7200RPM, there is about 25J of rotational kinetic energy, whose mass equivalent is 2.8x10^-13g (0.28 femtograms).

Assuming 200 electrons per NAND floating gate with 3bits/cell TLC on a 2TB SSD, there would be 5.3x10^14 electrons, weighing about 0.5 femtograms.

Proposed DNS RFC 8767: Serving Stale Data to Improve DNS Resiliency (2020)

antimatter15 — Tue, 21 Oct 2025 02:47:21 +0000

Article URL: https://datatracker.ietf.org/doc/html/rfc8767

Comments URL: https://news.ycombinator.com/item?id=45651921

Points: 3

# Comments: 0

A Cheeky Pint with Anthropic CEO Dario Amodei [video]

antimatter15 — Wed, 06 Aug 2025 16:03:25 +0000

Article URL: https://www.youtube.com/watch?v=GcqQ1ebBqkc

Comments URL: https://news.ycombinator.com/item?id=44813827

Points: 2

# Comments: 0

New comment by antimatter15 in "Up to date prices for LLM APIs all in one place"

antimatter15 — Fri, 25 Jul 2025 16:41:07 +0000

The `ccusage` npm package pulls prices and other information from LiteLLM which has a lot of diferent models: https://raw.githubusercontent.com/BerriAI/litellm/main/model...

Show HN: Vibechat – A chatroom for people bored waiting for Claude

antimatter15 — Mon, 07 Jul 2025 15:23:06 +0000

Article URL: https://github.com/antimatter15/vibechat

Comments URL: https://news.ycombinator.com/item?id=44491324

Points: 10

# Comments: 0

DumPy: NumPy except it's OK if you're dum

antimatter15 — Sat, 24 May 2025 01:05:47 +0000

Article URL: https://dynomight.net/dumpy/

Comments URL: https://news.ycombinator.com/item?id=44078019

Points: 28

# Comments: 1

New comment by antimatter15 in "A first successful factorization of RSA-2048 integer by D-Wave quantum computer"

antimatter15 — Sun, 11 May 2025 12:54:47 +0000

Note that this doesn't represent a general break of RSA-2048, and doesn't affect the security of RSA-2048 as it's used anywhere.

The paper only applies to "special integers" where the prime factors are known to only differ by two bits.

New comment by antimatter15 in "S1: A $6 R1 competitor?"

antimatter15 — Thu, 06 Feb 2025 20:20:59 +0000

It's hardly even illicit- at least in the United States, the output of an AI isn't copyrightable.

New comment by antimatter15 in "La Basilica Di San Pietro"

antimatter15 — Wed, 20 Nov 2024 19:40:53 +0000

Looking at the source code with web inspector it seems to be powered by 3D gaussian splatting and BabylonJS (https://doc.babylonjs.com/features/featuresDeepDive/mesh/gau...).

New comment by antimatter15 in "Ask HN: Fast data structures for disjoint intervals?"

antimatter15 — Tue, 23 Jul 2024 15:44:35 +0000

Maybe look into implicit in-order forests (https://thume.ca/2021/03/14/iforests/)

React 19 Breaks Async Composability

antimatter15 — Thu, 13 Jun 2024 21:59:03 +0000

Article URL: https://github.com/facebook/react/pull/26380

Comments URL: https://news.ycombinator.com/item?id=40675315

Points: 88

# Comments: 110

New comment by antimatter15 in "Financial Statement Analysis with Large Language Models"

antimatter15 — Fri, 24 May 2024 18:10:49 +0000

Figure 3 on p.40 of the paper seems to show that their LLM based model does not statistically significantly outperform a 3 layer neural network using 59 variables from 1989.

  This figure compares the prediction performance of GPT and quantitative models based on machine learning. Stepwise Logistic follows Ou and Penman (1989)’s structure with their 59 financial predictors. ANN is a three-layer artificial neural network model using the same set of variables as in Ou and Penman (1989). GPT (with CoT) provides the model with financial statement information and detailed chain-of-thought prompts. We report average accuracy (the percentage of correct predictions out of total predictions) for each method (left) and F1 score (right). We obtain bootstrapped standard errors by randomly sampling 1,000 observations 1,000 times and include 95% confidence intervals.

New comment by antimatter15 in "Humane AI – Pico Laser Projection – AI Twist on an Old Scam (2023)"

antimatter15 — Wed, 17 Apr 2024 15:00:53 +0000

For dimly-lit environments, the human eye's peak sensitivity for scoptopic vision is around 498nm (https://en.wikipedia.org/wiki/Scotopic_vision) which is blueish-green.

New comment by antimatter15 in "JetMoE: Reaching LLaMA2 performance with 0.1M dollars"

antimatter15 — Thu, 04 Apr 2024 19:26:33 +0000

It looks like Llama 2 7B took 184,320 A100-80GB GPU-hours to train[1]. This one says it used a 96×H100 GPU cluster for 2 weeks, for 32,256 hours. That's 17.5% of the number of hours, but H100s are faster than A100s [2] and FP16/bfloat16 performance is ~3x better.

If they had tried to replicate Llama 2 identically with their hardware setup, it'd cost a little bit less than twice their MoE model.

[1] https://github.com/meta-llama/llama/blob/main/MODEL_CARD.md#...

[2] https://blog.ori.co/choosing-between-nvidia-h100-vs-a100-per...

New comment by antimatter15 in "The Era of 1-bit LLMs: ternary parameters for cost-effective computing"

antimatter15 — Thu, 29 Feb 2024 03:10:47 +0000

They cite straight through estimators in the previous work with many of the same authors on (actual binary) BitNet

FCC Reaffirms Rejection of Nearly $900M Subsidy to Starlink [pdf]

antimatter15 — Tue, 12 Dec 2023 23:55:32 +0000

Article URL: https://docs.fcc.gov/public/attachments/DOC-399068A1.pdf

Comments URL: https://news.ycombinator.com/item?id=38620624

Points: 4

# Comments: 0

New comment by antimatter15 in "Phi-2: The surprising power of small language models"

antimatter15 — Tue, 12 Dec 2023 18:39:23 +0000

Looks like it is possible to download it locally, but as far as I can tell you have to manually copy all the various files from the Artifacts folder individually