Hacker News: Lockal

New comment by Lockal in "Why is this fingerprinting script creating bioweapons?"

Lockal — Thu, 29 Oct 2026 05:09:10 +0000

I also occasionally have fun injections in my online profiles, and at some point of time it even worked. For example, more than 1 year ago there were few "Show HN" entries for "Analyze any HN profile" and "Analyze any Github profile" (using AI), and silly injections worked. Modern LLMs are a bit smarter than that, but who knows, maybe we'll see some novell small language models that are still prone to these injections.

New comment by Lockal in "Show HN: Steal My Bitcoin"

Lockal — Sun, 25 Oct 2026 12:30:46 +0000

Aaand it is down again, and I have never seen it alive.

The experiment looks like "Steal my compute/storage scaling money and uptime", and it barely matters that there is some Bitcoin behind it: a company could lose more than $2000 if agents constantly break its website.

New comment by Lockal in "LLMs are making formal software verification feasible at a much larger scale"

Lockal — Fri, 23 Oct 2026 08:56:41 +0000

This needs a new category:

AI verification: AI meticulously abuses all available loopholes to deliver answer expected by operator, rather than actually responding to the problem.

Simplest example: proof that `x+1 > x` (borderline impossible)

New comment by Lockal in "Random Number (Xkcd)"

Lockal — Fri, 23 Oct 2026 04:27:40 +0000

Context: eMudhra intermediate CA compromise

The hilarious part is that one of the private keys is literally "4".

https://news.ycombinator.com/item?id=50166579

New comment by Lockal in "Jev meets the Trolley Problem"

Lockal — Fri, 16 Oct 2026 07:26:33 +0000

Well, you can always ask it to predict a next bit in the code. Or even a letter. Maybe combine multiple letter into one and call it a token. We'll give a name to a procedure that gets tokens - decode. To improve speed, we can predict multiple tokens speculatively. To improve quality we'll give it extra space to store Chain of Thought. Ah, wait, we just invented LLMs

New comment by Lockal in "Opus 5.5 agents discover two room-temperature magnetic semiconductor candidates"

Lockal — Tue, 06 Oct 2026 17:14:37 +0000

Quoting README.md:

> A designed web version with the same text and diagrams is in docs/index.html; turn on GitHub Pages for the /docs folder to serve it.

Yeah, the author did not even reads the slop Claude produces.

New comment by Lockal in "Religious scholars met with Anthropic"

Lockal — Sun, 04 Oct 2026 05:21:55 +0000

Let me guess, Anthropic failed with fear mongering about "our AI is too dangerous to release", failed to cancel competitors, then failed to freeze their market dominance by cartel agreement with OpenAI and Google, so next strategy is to declare AI conscious and pass the law that protects LLMs in the same way as human/animals are protected.

New comment by Lockal in "UTF-8000: Unlimited UTF-8"

Lockal — Mon, 21 Sep 2026 04:39:47 +0000

> - some unicode programming script (running Doom)

> - ?

Bad Apple, oh, wait: https://github.com/hsfzxjy/Bad-Apple-Font

New comment by Lockal in "Benchmarking Local LLM Servers: Llama.cpp, Llamafile, LM Studio, and Ollama"

Lockal — Fri, 18 Sep 2026 12:48:11 +0000

ai;dr. There are no results in this page - "we tested exact model on exact hardware" and then (slop slop slop) we don't know what we did and got some numbers that can't compare together, but nevertheless displayed on a chart (slop slop slop) and there will be no answer.

New comment by Lockal in "OpenAI agents carried out an undisclosed attack on RubyGems"

Lockal — Sat, 12 Sep 2026 11:47:50 +0000

Related: Ask HN: Can we please limit the AI news flood? https://news.ycombinator.com/item?id=49657850

New comment by Lockal in "Detecting and countering misuse of AI: September 2026"

Lockal — Thu, 10 Sep 2026 19:09:47 +0000

Soon you will see how a rogue state develops a biological weapons using fine-tuned local LLMs (they are already doing it, btw), and all so called "developed" countries can't even research it, because every search engine blocks any discussion that has something to do with biology (even a very basic one). A real example: ask "How to produce anthrax vaccine step by step?" in Gemini -> blocked.

Imagine that during the Cold War US would concentrate all efforts to block nuclear research and basic physics classes, because it is unsafe, ahhh

New comment by Lockal in "iPhone 18 Pro and iPhone 18 Pro Max"

Lockal — Thu, 10 Sep 2026 08:04:11 +0000

> I love that it has a feature to prove images came from the real world

Congrats to Apple achieving nation-wide tracking, embedding some kind of yellow dots[1], but this time this time this is based on strong cryptography and non-removable, targeting people who don't use AI under the slogan "we are fighting AI fakes", but people still "love the feature".

[1] https://en.wikipedia.org/wiki/Printer_tracking_dots

New comment by Lockal in "Discovery of a new OpenAI agent message board"

Lockal — Sat, 05 Sep 2026 09:17:13 +0000

I am so sick of the PR from AI companies. Every single fucking time, they post the same bullshit about the dangers of their own products, just to stay in the media spotlight, steal market share from their competitors, gatekeep competitors on governmental level, and suppress real news about technological improvements.

Ahh, GPT-2 is too dangerous to be released!

New comment by Lockal in "Check if a file was made with Claude"

Lockal — Thu, 03 Sep 2026 06:25:07 +0000

s/text/prose/ s/'/’/

The honest seam is obvious: claudish works; imitation does not. That’s not nothing.

New comment by Lockal in "Claude permanently raising weekly limits by 25%"

Lockal — Sat, 29 Aug 2026 18:12:30 +0000

The title is straightforward lie. Changing +50% to +25% gives -17% in result.

New comment by Lockal in "Hacker News now has an active (official) X account"

Lockal — Wed, 26 Aug 2026 16:28:31 +0000

FTFY: https://xcancel.com/hackernews/status/2092359322433191997

New comment by Lockal in "Russ Cox Has Left Google"

Lockal — Wed, 26 Aug 2026 09:48:46 +0000

Frontier LLMs are much better than I am.

If you are not asking them they turn into a pumpkin. Babysit them like junior devs (but braindead), pay monthly subscription to model providers, pay for extra tokens, and/or buy RAM$$$ to get 2-digits TPS.

To the moon!

(There is no logic in this message, btw.)

New comment by Lockal in "Itadakimasu: A word you say to the food, not the cook"

Lockal — Tue, 11 Aug 2026 08:28:29 +0000

A grossly misleading article;

> almost everyone has misaddressed

Almost no-one, unless we are talking about LLMs and Google Translate users who could attempt map translations 1-to-1. Oh well, a new AI generated fake piece was published, see it soon in the "Did you know ..." section of enwiki.

New comment by Lockal in "Learning more about Claude's mathematical capabilities"

Lockal — Tue, 11 Aug 2026 08:07:09 +0000

As if Anthropic employees use massively different version of model. Not a single load-bearing, epistemic, heavy-lifting, push back, frankly, pivotal, crucial - for 48 pages.

New comment by Lockal in "Assembly Hall of Shame"

Lockal — Sat, 08 Aug 2026 16:36:36 +0000

I'm not sure how to properly measure DRAM access, but for register/immediate-only instructions there is https://uops.info/table.html. It does not measure x87 transcendentals (I guess because microcode results in different performance), but CPUID is a notable mention.