Hacker News: crimsoneer

New comment by crimsoneer in "Hubble: Open-source notetaking app for you and your agents"

crimsoneer — Wed, 29 Jul 2026 10:59:21 +0000

Sorry, but wtf is this and why is it on the front-page??

"Agent ready"? Eg, it uses markdown like... everything else? There is literally no information here.

New comment by crimsoneer in "Tangleflow: Converts GitHub Actions workflows to tangled workflows and back"

crimsoneer — Fri, 24 Jul 2026 12:46:50 +0000

It's quite a cool technical concept, and I really like their design sensibility, but don't know anybody using it "in prod".

New comment by crimsoneer in "Ask HN: Does anyone let AI agents play games just for fun?"

crimsoneer — Mon, 13 Jul 2026 12:25:51 +0000

Yes!

I spent ages watches them play Risk. It was fun and deeply silly: https://andreasthinks.me/posts/ai-at-play/

I've now got them playing Blood Bown(ish), and they're bad: https://ai-at-play.online/

New comment by crimsoneer in "Al Vigier: Canada's AI strategy shouldn't include secret Palantir bills"

crimsoneer — Mon, 06 Jul 2026 19:32:53 +0000

The idea we're going to stop people like special forces and intelligence buying the software they want to run on on prem, air gapped stacks that they control in the name of sovereignty is silly.

New comment by crimsoneer in "Real-time map of Great Britain's rail network"

crimsoneer — Mon, 06 Jul 2026 12:05:33 +0000

I suspect it's linked to this which is the latest example being used.

https://www.ndtv.com/world-news/uk-man-was-assaulted-by-atta...

New comment by crimsoneer in "Real-time map of Great Britain's rail network"

crimsoneer — Mon, 06 Jul 2026 11:07:14 +0000

This feels like a weird takeway to take from this!

Out of curiousity, where are you from? I find it interesting where only people who get so exercised about the UK get their info from...

New comment by crimsoneer in "Bring back crappy forums"

crimsoneer — Thu, 02 Jul 2026 09:18:50 +0000

It's worth reflecting on the fact nobody has ever been arrested for running a forum under the online safety act. One forum, which was specifically dedicated to helping people commit suicide, was fined.

https://www.theguardian.com/technology/2026/feb/27/suicide-f...

New comment by crimsoneer in "Bring back crappy forums"

crimsoneer — Thu, 02 Jul 2026 09:10:35 +0000

Why? You are significantly more likely to be struck by lightning when out for a walk then prosecuted for running a forum under the online safety act.

The act and guidance are very, very clear that your safeguards have to be proportionate to your scale and audience. If you're running a small forum, a quick "I sometimes check the posts and make sure nobody is using it to groom children or exchange child porn" is just fine.

New comment by crimsoneer in "Good results fine tuning a local LLM like Qwen 3:0.6B to categorize questions"

crimsoneer — Mon, 22 Jun 2026 08:23:28 +0000

Tangentially related, but the UK Gov Incubator for AI has quite a nifty LLM driven classification pipeline for survey answers.

https://github.com/i-dot-ai/consult

New comment by crimsoneer in "GLM 5.2 vs. Opus"

crimsoneer — Mon, 22 Jun 2026 08:20:29 +0000

Practically nobody.

New comment by crimsoneer in "Under-16s to be banned from social media, Starmer announces"

crimsoneer — Mon, 15 Jun 2026 08:08:59 +0000

This specifically doesn't ID anybody. It's not using the upcoming digital ID, it's just going to use the existing age authentication - eg, you just need to show your company is doing some sort of "serious" age verification (not just enter your DOB here) rather than actually checking identity.

It's bad, shallow populism, but it's not some nefarious surveillance plan, as much as the yanks are going to paint it that way.

New comment by crimsoneer in "Let's talk about EU Sovereignty (2025)"

crimsoneer — Sat, 30 May 2026 20:07:55 +0000

Bring back the minitel.

New comment by crimsoneer in "Notes from the Mistral AI Now Summit"

crimsoneer — Fri, 29 May 2026 19:20:12 +0000

Based on what? Is there any evidence of risk at all?

New comment by crimsoneer in "The Structural Barriers to AI Lawyers"

crimsoneer — Wed, 27 May 2026 08:00:51 +0000

Fair, most countries that follow the common law judicial system I should have said.

New comment by crimsoneer in "The Structural Barriers to AI Lawyers"

crimsoneer — Wed, 27 May 2026 07:55:12 +0000

This isn't really true though. This is how the law used to work, until people did the research and discovered it let to absolutely loads of mad variation in outcomes, with people with similar offences getting totally different sentences based on random luck. Hence most countries not have pretty strict sentencing guidelines, with a bit of space for judgement on top (despite a lot of protesting from judges).

https://www.ubs.com/global/en/our-firm/what-we-do/our-brand/...

You should be able to predict the outcome of a court case if you have all the facts available. That's what fairness means.

New comment by crimsoneer in "Outsourcing plus local AI will soon become more economical vs. frontier labs"

crimsoneer — Tue, 26 May 2026 14:26:55 +0000

I think this is a compelling argument, but I think 2 issues:

1. I remain unconvinced LocalAI can work well for majority of businesses. It looks vaguely comparable on benchmarks, but it tends to be fragile and a lot of management overhead in reality.

2. Similarly, while Deepseek is comparable to Opus/Codex on benchmarks, for agentic work at scale I definitely notice the difference. That's not to say it's not economical, just that I definitely miss the big boys when I swap.

I kind of wish this was true, because the UK would be in a great place to compete with the US. But somehow people are happy to pay 3x the salary for an engineer in SF.

New comment by crimsoneer in "Show HN: I made a tactical map-based WWII submarine simulator (public beta)"

crimsoneer — Thu, 21 May 2026 07:08:26 +0000

Is that comic sans or am I having a moment

New comment by crimsoneer in "London Police Deploy Facial Recognition at Protest for First Time"

crimsoneer — Sat, 16 May 2026 04:30:17 +0000

It's worth recognising that unlike oyster cards, LFR in it's currently form doesn't store any data. It looks at faces, compares them to currently wanted offenders, and if it matches, alerts the officers operating that camera "live". Nothing is stored or processed beyond that

New comment by crimsoneer in "London Police Deploy Facial Recognition at Protest for First Time"

crimsoneer — Sat, 16 May 2026 04:27:04 +0000

This is very naive.

1. The odds of the actual bike itself being covered by CCTV during the theft is pretty slim. Nearby? Sure, but probably not recording the offence. Then unless you're got a precise time window/know where the thief went, you're stuck watching hours of footage hoping to spot the right bike.

2. Even if you do get a clear, high quality facial picture of the thief, you have no magic way of figuring out who that is. You essentially email it to all the local cops and hope someone's recognises them.

The result is your bike theft turns into quite a big investigate, with a sub 5% likelihood of a position outcome.

New comment by crimsoneer in "European Money Pours into Palantir"

crimsoneer — Wed, 13 May 2026 00:37:20 +0000

You do realise saying this about a publicly traded company without any evidence is just full on qanon level conspiracy theory right?