Hacker News: onoesworkacct

New comment by onoesworkacct in "I accidentally turned LLM memory into program analysis"

onoesworkacct — Sat, 29 Aug 2026 09:37:47 +0000

yooo.. this is exactly what I wanted to do... I'm so happy because (a) I hadn't figured out how to do it (b) it seemed kind of difficult in general and (c) now I don't have to, lol.

New comment by onoesworkacct in "Review: Chuwi's $449 Unibook laptop is a funhouse-mirror MacBook Neo"

onoesworkacct — Sat, 29 Aug 2026 09:21:35 +0000

Using a laptop outside is fantastic. I love to sit on my balcony and work there instead of inside. My work laptop is an HP with zero ability to do this.

New comment by onoesworkacct in "Microduck"

onoesworkacct — Fri, 28 Aug 2026 00:19:18 +0000

IMO with that definition of "toy" you're overselling the value of performing useful work and underselling the value of farting about.

New comment by onoesworkacct in "Vomit: Clean up Claude 5's token output with a separate LLM"

onoesworkacct — Sat, 22 Aug 2026 00:16:37 +0000

That project looks needlessly complex, you can just spawn a Haiku subagent in the session and it does the job.

New comment by onoesworkacct in "Vomit: Clean up Claude 5's token output with a separate LLM"

onoesworkacct — Sat, 22 Aug 2026 00:15:41 +0000

This doesn't work for me. The output gets messed up when the model is overloaded, no matter what. The cognitive load is too much if it has to optimise for readability at the same time.. or it's not trained to be able to do that.

I usually get output at the end of a long task. At that point I'm going to use a subagent with no context to reword it. Haiku does a great job with writing style so I've been using that, the main agent will fix any inaccuracies.

New comment by onoesworkacct in "Georgia police officers fired after Flock camera misuse"

onoesworkacct — Mon, 10 Aug 2026 00:25:11 +0000

Who's more effective at protesting, USA or France?

And, which country routinely sets a bunch of shit on fire when they need to?

That's right.

New comment by onoesworkacct in "The Hacker's Renaissance (2025)"

onoesworkacct — Sun, 09 Aug 2026 23:16:58 +0000

It was written by ChatGPT which would avoid examples that are explicitly illegal, and would stop short of advocating things that contravene ToS agreements.

Sad.

Who is in charge of this shit now?

New comment by onoesworkacct in "DeepSeek V4 Flash on a Single AMD MI300X"

onoesworkacct — Wed, 05 Aug 2026 05:36:34 +0000

Demand is money.

No bubble, where money?

New comment by onoesworkacct in "The Warp Agent CLI"

onoesworkacct — Wed, 05 Aug 2026 05:10:39 +0000

just... ssh into them and then run the agent?

New comment by onoesworkacct in ""An AI Job Apocalypse?" – Goldman Sachs Report [pdf]"

onoesworkacct — Fri, 03 Jul 2026 01:54:35 +0000

By "the alternative" are you describing our present reality? Because it sounds about right

New comment by onoesworkacct in "GLM 5.2 beats Claude in our benchmarks"

onoesworkacct — Mon, 29 Jun 2026 04:15:10 +0000

could mimo have scraped the mythos findings already? it's very recent

New comment by onoesworkacct in "5k menus from the New York Public Library’s Buttolph Collection (1880-1920)"

onoesworkacct — Mon, 29 Jun 2026 03:09:31 +0000

They're a mainstay at my local Japanese and Korean BBQ restaurants.

New comment by onoesworkacct in "GLM 5.2 beats Claude in our benchmarks"

onoesworkacct — Mon, 29 Jun 2026 01:45:18 +0000

The rumour is that it's trained on Opus, but who knows

New comment by onoesworkacct in "Has AI already killed self-help nonfiction books?"

onoesworkacct — Wed, 17 Jun 2026 04:39:14 +0000

I haven't read it yet, but reading it is not going to change your personality. Personality defines how you receive it.

If you change your entire personality based on a self-help book.. that probably says a lot about your personality.

And anyway twin studies make the hardware seem more impactful than the software in many ways..

New comment by onoesworkacct in "If Claude Fable stops helping you, you'll never know"

onoesworkacct — Wed, 10 Jun 2026 00:52:31 +0000

most people don't wanna do that. there are plenty of people who would infect people with crypto botnets

New comment by onoesworkacct in "Show HN: Semble – Code search for agents that uses 98% fewer tokens than grep"

onoesworkacct — Mon, 18 May 2026 00:52:37 +0000

fantastic token savings and performance... but unlike grep it's probabilistic search on search terms.

is that an issue? the tiny model might not surface something important

New comment by onoesworkacct in "Trusted access for the next era of cyber defense"

onoesworkacct — Tue, 14 Apr 2026 23:47:36 +0000

It was. The internet really has been filled with abject shit and social media is bots talking to bots.

New comment by onoesworkacct in "Anthropic downgraded cache TTL on March 6th"

onoesworkacct — Mon, 13 Apr 2026 07:05:07 +0000

give it a skill that runs a timer in the background and every 4.5 minutes says "ping? pong!"

New comment by onoesworkacct in "Issue: Claude Code is unusable for complex engineering tasks with Feb updates"

onoesworkacct — Tue, 07 Apr 2026 02:09:07 +0000

This kind of thing is harder for regular end-users to understand following the change removing reasoning details.

New comment by onoesworkacct in "Iran strikes leave Amazon availability zones "hard down" in Bahrain and Dubai"

onoesworkacct — Sat, 04 Apr 2026 05:22:35 +0000

you're going to save money by having your own physical servers?