Hacker News: skerit

New comment by skerit in "Prompting Claude Opus 5.5"

skerit — Mon, 28 Sep 2026 10:15:34 +0000

Opus 5.5 is just too eager in everything it does. The only good thing about this is that it's most often doing something correct.

New comment by skerit in "Prompting Claude Opus 5.5"

skerit — Mon, 28 Sep 2026 10:13:49 +0000

I've switching from only using markdown in my prompts to using XML tags this year too. It's not only easy for the model to see when something ends, it's quite useful for me too.

New comment by skerit in "Prompting Claude Opus 5.5"

skerit — Mon, 28 Sep 2026 10:11:41 +0000

> In Anthropic's testing, at its default "medium" effort the model matched or beat Claude Opus 5 at "high" effort on such tasks, in fewer steps and with fewer tokens

Opus 5.5 has been amazing, but I'm confused by how this is worded. It "matched or beat" Opus 5? There is no matching. There is only surpassing. By miles. Like Opus 5 was the biggest disappointment of the year. Opus 5.5 is even better than Fable. I do not understand why they're not acknowledging it for the leap that it is?

New comment by skerit in "Claude Opus 5.5"

skerit — Tue, 22 Sep 2026 18:12:30 +0000

Retiring the Terra tier? Their space-inspired lineup has only been out for 2 months, they're already messing with it?

New comment by skerit in "Claude Opus 5.5"

skerit — Tue, 22 Sep 2026 18:07:43 +0000

I like the Pelican test. And I agree this pelican looks very boring.

New comment by skerit in "Claude Opus 5.5"

skerit — Tue, 22 Sep 2026 18:05:00 +0000

I'm good with the odds on those 2 scenarios. I believe humans could kill all life on earth without AI anyway.

New comment by skerit in "Btrfs/ZFS/bcachefs under workloads classic benchmarks skip"

skerit — Sat, 19 Sep 2026 18:49:06 +0000

Oh, so bcachefs is doing pretty well.

New comment by skerit in "Introducing System One Models and Jev"

skerit — Tue, 15 Sep 2026 20:17:50 +0000

So in theory you could feed it incomplete text, and then ask it for the probabilities of what the next character could be?

New comment by skerit in "JetKVM Mini"

skerit — Sun, 13 Sep 2026 08:16:16 +0000

At least you got them. I'm still at the "We will send you an email to do a final shipment confirmation once ready" stage.

New comment by skerit in "Altman tells staff OpenAI is open to slowing AI development"

skerit — Sat, 12 Sep 2026 12:45:11 +0000

All they're saying to the other players is: let's lock it down and slice up the market between us.

Claiming Skynet is just around the corner is a great way to get the public to believe only these huge companies can harness AI, and how all open-source models should be outlawed.

New comment by skerit in "216M Spy TVs – The LG Smart TV Problem [video]"

skerit — Mon, 07 Sep 2026 13:35:56 +0000

I never understood why any of the smartness had to be built into the TV in the first place. Sure, it's useful for the first year, and then the seriously underpowered hardware they installed into it will have trouble with just about anything.

I bought a Philips Ambilight OLED TV probably over 8 years now. Brilliant tv, great quality, I still see no reason to replace it at all. But its built in AndroidTV is garbage.

New comment by skerit in "Ask HN: Who is using MCP in production?"

skerit — Fri, 04 Sep 2026 10:18:22 +0000

I use MCPs all the time, but just like in your case: they're my own MCP servers.

They do exactly what I want and how I want it. They expose just the information and the features I want the LLM to have. And just like in your case, it combines multiple sources of information into 1 response.

New comment by skerit in "Microduck"

skerit — Fri, 28 Aug 2026 08:38:55 +0000

It is a bit weird. As someone from Belgium (we also use our own variety of Azerty) you are very frequently reminded that you're on a weird keyboard. At least for someone who games, you wouldn't make ZQSD the default. It's the first thing you do in every game, change WASD to ZQSD

New comment by skerit in "Don't Wordle"

skerit — Wed, 26 Aug 2026 10:23:28 +0000

So including me, you're like the fourth person in this section to get it on the second turn. That's peculiar.

New comment by skerit in "Don't Wordle"

skerit — Wed, 26 Aug 2026 10:08:36 +0000

What the hell, same here. My initial guess was "CELLO", from there on I had 32 other possible words and somehow I got the right one?

New comment by skerit in "Anthropic's best AI model struggles to attract users as cheaper tools thrive"

skerit — Mon, 24 Aug 2026 11:40:33 +0000

> The individual plans are heavily subsidized

Not this again. There is not a single shred of evidence for this. In fact multiple times this year alone, people from Anthropic have said that inference and deployed models have positive margins. The big bucks are always being spent on training the next model.

New comment by skerit in "A week of using Codex more than Claude"

skerit — Sun, 23 Aug 2026 12:09:45 +0000

> A week of using Codex more than Claude

Which Claude is this about? Sonnet, Opus, Fable? All totally different beasts.

New comment by skerit in "Don't Paste the AI, please"

skerit — Thu, 20 Aug 2026 12:02:49 +0000

I agree that "The person on the other side has the same tools you do" might be technically true, I have myself encountered it often enough that they don't even use it at all.

People don't always want MY take on it, they just don't want to bother finding the solution on their own at all.

So in that case replying with a verbatim copy-paste of a Claude response has the same energy as a lmgtfy link.

New comment by skerit in "Claude Code weekly limits reduce by a third tomorrow"

skerit — Tue, 18 Aug 2026 18:25:22 +0000

Meanwhile every time their servers go down, your caches go down with it and you can pay for your entire context again. This sucks even harder when you like to use parallel subagents.

New comment by skerit in "Degraded performance for multiple models"

skerit — Tue, 18 Aug 2026 17:15:16 +0000

It's been a while since the last reset. I think we're due one. Though I would prefer they just extend the +50% usage limit forever, it's been so long I can not imagine lossing a third of my current usage.