Hacker News: semiquaver

New comment by semiquaver in "200 Milliseconds"

semiquaver — Mon, 03 Aug 2026 23:02:41 +0000

  > A problem repeatedly occurred on "https://200ms.thenodebook.com/#act-O-prologue”

I find this usually is caused by being too clever with layouts.

New comment by semiquaver in "OpenAI’s amazing — but vastly oversold — new model Astra"

semiquaver — Mon, 03 Aug 2026 00:03:07 +0000

The cope is tangible. Article is drenched in flop sweat to a remarkable extent, even for Mr. Marcus. I cannot see the goal posts, they have been moved so much.

New comment by semiquaver in "Deflock Casa Grande"

semiquaver — Tue, 28 Jul 2026 19:27:54 +0000

Nothing about this site screams AI to me. The site does its job well. Not sure what you are complaining about.

New comment by semiquaver in "LLM Usage in Debian: Three Proposals"

semiquaver — Sun, 26 Jul 2026 16:48:53 +0000

It’s not sampling anything though. At each inference step the fundamental output of the process can be interpreted as a probability for each possible token how likely it is. Sampling would imply otherwise but the probabilities are the output.

New comment by semiquaver in "London Gatwick has launched a robotic airport parking service"

semiquaver — Sun, 26 Jul 2026 16:40:16 +0000

  > (probably, given recent AI advances) an incredible time to market

There’s nothing whatsoever to indicate that this has anything at all to do with LLMs.

New comment by semiquaver in "Inflect-Micro-v2: complete voice in 9.36M parameters"

semiquaver — Sun, 26 Jul 2026 16:33:00 +0000

When would “text-to-waveform speech synthesis” ever imply speech to text?

New comment by semiquaver in "Claude Cookbook"

semiquaver — Fri, 24 Jul 2026 18:33:23 +0000

Ok, if there was chiptune music I think I _would_ like the after shots better.

New comment by semiquaver in "Claude Cookbook"

semiquaver — Fri, 24 Jul 2026 14:18:16 +0000

The before and after images in “prompting for frontend aesthetics” are hilarious. Did no one look at these to make sure the skill actually improved the design?

https://platform.claude.com/cookbook/coding-prompting-for-fr...

New comment by semiquaver in "Does creatine make you smarter?"

semiquaver — Thu, 23 Jul 2026 00:26:54 +0000

Do you regularly measure your own blood pressure on a daily basis? I’d imagine if this were common most people wouldn’t know.

New comment by semiquaver in "Any text-to-SQL benchmark should address difficulties of real-world data stores"

semiquaver — Wed, 22 Jul 2026 22:38:23 +0000

  > It is widely known that an LLM can only find data it has seen before.

What does this mean? It’s obviously not literally true, so what is the author trying to convey?

New comment by semiquaver in "“We have information that Moonshot distilled Fable for the development of K3”"

semiquaver — Wed, 22 Jul 2026 21:09:04 +0000

This misstates the small number of legal opinions and orders on this topic, none of which form binding precedent outside the districts where the cases happened. So even if a court had found that “LLM output is public domain” (none did) that wouldn’t make it “the law” until it went up the appellate system and was upheld.

Our current laws simply weren’t built for this and I expect the legal status of LLM output is not going to be resolved until Congress actually legislates on this topic.

New comment by semiquaver in "Nobody knows what a used GPU cluster is worth"

semiquaver — Wed, 22 Jul 2026 17:55:10 +0000

Maybe this article has something useful to say but the painfully LLM-generated prose is too distracting to make it evident.

New comment by semiquaver in "'VPNs are lawful technical tools,' says EU Court in landmark copyright ruling"

semiquaver — Wed, 22 Jul 2026 13:46:02 +0000

  > People do care about income streams for their descendants or charitable organizations

Society created the concept of copyright and intellectual property for a reason and it is emphatically not the protection of income streams after you are dead. Whether a person exists who cares about a thing is not a reason to preserve it.

New comment by semiquaver in "OpenAI and Hugging Face address security incident during model evaluation"

semiquaver — Tue, 21 Jul 2026 21:55:12 +0000

What on earth is the liability situation for these models? If OpenAI has a monster in a lab that is doing real world monetary harm to other companies, could those parties sue for damages over it? Or could OAI be charged criminally for the many varied CFAA violations which definitely happened here? I get that in this case that wont happen but it’s only a matter of time before these questions are no longer hypothetical.

New comment by semiquaver in "Apple Defeats Liability for Not Scanning iCloud for CSAM"

semiquaver — Tue, 21 Jul 2026 15:43:08 +0000

The judge’s dicta about protecting children in her pro-privacy ruling upholding existing law “tips their hand” that they are somehow part of a global conspiracy to eliminate privacy?

I think your conspiracy theory needs work, to be perfectly honest with you.

New comment by semiquaver in "Transcribe.cpp"

semiquaver — Sun, 19 Jul 2026 02:52:12 +0000

That would be Diarize.cpp, not Transcribe.cpp.

New comment by semiquaver in "GPT-5.6 used a prompt to close a 30-year gap in convex optimization"

semiquaver — Sat, 18 Jul 2026 18:57:06 +0000

I would not characterize it thus.

New comment by semiquaver in "GPT-5.6 used a prompt to close a 30-year gap in convex optimization"

semiquaver — Sat, 18 Jul 2026 17:42:50 +0000

Thanks for posting this comment, it makes me proud of myself to be able to partially comprehend the comment :)

New comment by semiquaver in "GPT-5.6 used a prompt to close a 30-year gap in convex optimization"

semiquaver — Sat, 18 Jul 2026 15:18:12 +0000

  > you can't have one LLM to read your mind to prompt another LLM

I’m excited to inform you that we as a species have developed a particularly useful facility known as Language which these LLM tools are evidently rather handy at wielding. This facility is particularly useful in this context when it takes the form of “dialog” or “questioning”, which can be used to propagate abstract ideas by means of mutually-feedback-guided-iterative-Language-use-turns, or more concisely, “conversation.”

One might even say that this remarkable facility can be used to “read” the ideas from one entity’s mind, such that after sufficient dialog the second entity obtains a (possibly lossy, but there are mitigations for this) copy of the ideas of the first. You might further be surprised to learn that this sort of idea-transfer business using language has already been happening in our society and species for quite some time indeed.

New comment by semiquaver in "GPT-5.6 used a prompt to close a 30-year gap in convex optimization"

semiquaver — Sat, 18 Jul 2026 13:46:03 +0000

You’re at least 18 months out of date claiming that prompting will be the new hot skill. Turns out LLMs are also good at prompting other LLMs.