Hacker News: artemisart

New comment by artemisart in "Opus 5 is currently #1 on Artificial Analysis Intelligence Leaderboard"

artemisart — Sat, 25 Jul 2026 14:33:07 +0000

Yes same website https://artificialanalysis.ai/models#intelligence-comparison... but they don't have graphs for the individual benchmarks sadly.

New comment by artemisart in "Claude Opus 5"

artemisart — Fri, 24 Jul 2026 19:39:58 +0000

No it's a mean of 5 runs.

> We report FrontierCode’s overall score, a composite measure that grades each patch on blocking functional criteria (held-out unit tests) together with weighted code-quality rubric criteria, as mean@5.

They don't explain more in the system card, I guess higher effort levels could loose points on the code quality / scope / style / maintainability stuff?

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

artemisart — Fri, 17 Jul 2026 00:12:13 +0000

Official API doc says only max effort level is currently supported. https://platform.kimi.ai/docs/guide/kimi-k3-quickstart#think...

New comment by artemisart in "Claude Fable is relentlessly proactive"

artemisart — Fri, 12 Jun 2026 17:48:20 +0000

bwrap is builtin in claude too, activate with /sandbox command.

New comment by artemisart in "MAI-Code-1-Flash"

artemisart — Tue, 02 Jun 2026 20:47:07 +0000

Hill climbing doesn't mean much but absolutely doesn't imply they cheat on benchmarks. They have more details here https://microsoft.ai/news/introducing-mai-thinking-1/ it seems to be "RL on everything".

New comment by artemisart in "NSA is using Anthropic's Mythos despite blacklist"

artemisart — Mon, 20 Apr 2026 18:36:17 +0000

Every US intelligence org probably has at least API access, but anything outside of the US? No chance.

New comment by artemisart in "OpenZL: An open source format-aware compression framework"

artemisart — Mon, 06 Oct 2025 21:16:36 +0000

I may be misunderstanding the question but that should be just decompressing gzip & compressing with something better like zstd (and saving the gzip options to compress it back), however it won't avoid compressing and decompressing gzip.

New comment by artemisart in "GPT-5-Codex"

artemisart — Mon, 15 Sep 2025 23:24:15 +0000

Does refactoring mean moving things around for people? Why don't you use your IDE for this, it already handles fixing imports (or use find-replace) and it's faster and deterministic.

New comment by artemisart in "DuckDB NPM packages 1.3.3 and 1.29.2 compromised with malware"

artemisart — Wed, 10 Sep 2025 01:32:25 +0000

Do you know about other security issues? If it's only about curl | sh it really isn't a problem, if the same website showed you a hash to check the file then the hash would be compromised at the same time as the file, and with a package manager you still end up executing code from the author that is free to download and execute anything else. Most package managers don't add security.

New comment by artemisart in "DuckDB NPM packages 1.3.3 and 1.29.2 compromised with malware"

artemisart — Wed, 10 Sep 2025 01:15:11 +0000

Why should we expect companies to be able to reuse the correct token if they can't coordinate on using a single domain in the first place?

New comment by artemisart in "Mistral raises 1.7B€, partners with ASML"

artemisart — Tue, 09 Sep 2025 23:34:50 +0000

Yes for parakeet, but only comparing benchmark results for canary. Whisper also has severe hallucinations on silence and noise and WhisperX helps a lot, it adds voice activity detection i.e. a model to detect when someone speaks, to filter the input before running whisper. https://github.com/m-bain/whisperX

New comment by artemisart in "Mistral raises 1.7B€, partners with ASML"

artemisart — Tue, 09 Sep 2025 10:57:25 +0000

Nvidia parakeet and canary are better and faster, here is a leaderboard: https://huggingface.co/spaces/hf-audio/open_asr_leaderboard

New comment by artemisart in "Liquid Glass? That's what your M4 CPU is for"

artemisart — Thu, 04 Sep 2025 11:18:46 +0000

No, you never compute individual pixels because you never need to, and it's always faster to it in bulk (vectorization, memory access...) and so over an area you take the same number of pixels as input (or a little bit more with padding) and the blur will only increase significantly the compute.

New comment by artemisart in "What's New with Firefox 142"

artemisart — Tue, 02 Sep 2025 11:42:13 +0000

This seems to be exclusive to Safari, I can't get it to work in Chrome either (and didn't know about the feature before right now, the discoverability is terrible).

New comment by artemisart in "Nvidia DGX Spark"

artemisart — Thu, 28 Aug 2025 11:51:22 +0000

I don't understand what's not optimized on 5090. If we're comparing with Apple chips or AMD Strix Halo yes you will have very different hardware + software support, no FP4 etc. but here everything is CUDA, Blackwell vs Blackwell, same FP4 structured sparsity, so I don't get how it would be honest to compare a quantized FP4 model on Spark with an unoptimized FP16 model on a 5090 ?

New comment by artemisart in "Nvidia DGX Spark"

artemisart — Thu, 28 Aug 2025 10:41:35 +0000

Ok then just to clarify: you can fit 4x larger models on the Spark vs 5090, not 17x.

New comment by artemisart in "Nvidia DGX Spark"

artemisart — Thu, 28 Aug 2025 10:10:20 +0000

That's very true and what's segmenting the market, but I don't understand why you're saying the 5090 supports only 12B model when it can go up to 50-60B (= a bit less than 64B to leave room for inference) as it supports FP4 as well.

New comment by artemisart in "YouTube made AI enhancements to videos without warning or permission"

artemisart — Mon, 25 Aug 2025 10:51:38 +0000

The economics don't make sense, each video is stored ~ once (+ replication etc. but let's say O(1)) but viewed n times, so server-side upscaling on the fly is way too costly and currently not good enough client-side.

New comment by artemisart in "YouTube made AI enhancements to videos without warning or permission"

artemisart — Mon, 25 Aug 2025 10:47:40 +0000

> Then, once that is perfected, they will offer famous content creators the chance to sell their "image" to other creators, so less popular underpaid creators can record videos and change their appearance to those of famous ones, making each content creator a brand to be sold.

I'm frightened by how realistic this sounds.

New comment by artemisart in "Optician Sans – A free font based on historical eye charts and optotypes"

artemisart — Wed, 30 Jul 2025 22:01:00 +0000

Yes I don't understand how they can claim it's optimized for legibility when the base font does the inverse.