Hacker News: czk

New comment by czk in "GPT-5.5 hallucinates 3x more than MIT-licensed GLM-5.2"

czk — Sat, 20 Jun 2026 14:57:11 +0000

if you're benchmaxxing then maybe bigger doesnt always mean better, but for general intelligence and big model smell, that couldn't be further from the truth

the oss models are impressive but it's pretty clear how quickly they fall off when you try to use them outside of a narrow set of problems they benchmarked well on when compared to opus/5.5

New comment by czk in "HERMES.md in commit messages causes requests to route to extra usage billing"

czk — Wed, 29 Apr 2026 21:06:07 +0000

There was also a bug where you could cancel the subscription via the iOS app store and if you never opened the iOS claude app again, you'd keep the subscription forever and could use claude via the web, without paying.

Also when they added extra credits to everyone as an apology I was able to click the claim button multiple times and I got up to $400 in credits. Eventually a day later this dropped to $200 and then a few days later, $100 where it sits today.

New comment by czk in "DeepSeek v4"

czk — Fri, 24 Apr 2026 19:52:48 +0000

They mention it uses MXFP4 quant which is a blackwell capability but it looks like this is also supported by ascend 950 series according to marketing material

New comment by czk in "OpenAI releases GPT-5.5 and GPT-5.5 Pro in the API"

czk — Fri, 24 Apr 2026 19:47:09 +0000

i wonder if they put an older cutoff date into the prompt intentionally so that when asked on more current events it leans towards tool calls / web searches for tuning

New comment by czk in "OpenAI releases GPT-5.5 and GPT-5.5 Pro in the API"

czk — Fri, 24 Apr 2026 19:44:26 +0000

the model obviously knows things after the reported date but its just curious that it reports that date consistently

could be they do it intentionally to encourage more tool calls/searches or for tuning reasons

New comment by czk in "OpenAI releases GPT-5.5 and GPT-5.5 Pro in the API"

czk — Fri, 24 Apr 2026 19:40:28 +0000

with thinking off and tools disabled:

  Donald Trump won the 2024 U.S. presidential election.

New comment by czk in "OpenAI releases GPT-5.5 and GPT-5.5 Pro in the API"

czk — Fri, 24 Apr 2026 19:40:06 +0000

you cant but its pretty reproducible across api and codex and other agents so i just thought it was odd. full text it gives:

   Knowledge cutoff: 2024-06
   Current date: 2026-04-24

   You are an AI assistant accessed via an API.

   # Desired oververbosity for the final answer (not analysis): 5
   An oververbosity of 1 means the model should respond using only the minimal content necessary to satisfy the request, using
 concise phrasing and avoiding extra detail or explanation."
   An oververbosity of 10 means the model should provide maximally detailed, thorough responses with context, explanations, and
 possibly multiple examples."
   The desired oververbosity should be treated only as a *default*. Defer to any user or developer requirements regarding
 response length, if present.

New comment by czk in "OpenAI releases GPT-5.5 and GPT-5.5 Pro in the API"

czk — Fri, 24 Apr 2026 19:16:44 +0000

API page lists the knowledge cutoff as Dec 01, 2025 but when prompting the model it says June 2024.

   Knowledge cutoff: 2024-06
   Current date: 2026-04-24

   You are an AI assistant accessed via an API.

New comment by czk in "Qwen3.6-27B: Flagship-Level Coding in a 27B Dense Model"

czk — Wed, 22 Apr 2026 19:30:59 +0000

Memory bandwidth is the biggest L on the dgx spark, it’s half my MacBook from 2023 and that’s the biggest tok/sec bottleneck

New comment by czk in "The beginning of scarcity in AI"

czk — Thu, 16 Apr 2026 21:53:28 +0000

"adaptive" thinking

New comment by czk in "Claude Opus 4.7"

czk — Thu, 16 Apr 2026 20:32:04 +0000

show us the benchmarks with "adaptive thinking" turned on

New comment by czk in "Darkbloom – Private inference on idle Macs"

czk — Thu, 16 Apr 2026 17:12:09 +0000

the MDM profile requirement is suspect though I get why they are doing it. but it doesn't inspire confidence to see that their profile is unsigned and still using the default micromdn scep challenge...

New comment by czk in "Apple Silicon and Virtual Machines: Beating the 2 VM Limit (2023)"

czk — Sat, 11 Apr 2026 21:42:38 +0000

it should, lume is a thin wrapper around Apple's Virtualization.framework as i understand it

New comment by czk in "Apple Silicon and Virtual Machines: Beating the 2 VM Limit (2023)"

czk — Sat, 11 Apr 2026 21:38:33 +0000

starting with M3+ you can use Hypervisor.framework/Virtualization.framework to spin up nested VMs.

it would be amusing if that bypassed the limit.

New comment by czk in "Claude now creates interactive charts, diagrams and visualizations"

czk — Thu, 12 Mar 2026 17:17:11 +0000

I tried the periodic table in their examples using sonnet 4.6 on the $20/mo plan. After a few minutes Claude told me it reached the max message length and bailed. I pressed continue and eventually it generated the table, but it wasn't inline, it was a jsx artifact, and I've now hit my daily usage limit.

New comment by czk in "GPT‑5.3 Instant"

czk — Tue, 03 Mar 2026 23:46:35 +0000

claude models with 'extended thinking' toggled answer very quickly and the quality of the answer is far ahead of what gpt 5.2 'instant' provides. i wont even bother using the non-thinking version of chatgpt because the quality of the answers is awful and usually incorrect.

New comment by czk in "Claude is a space to think"

czk — Wed, 04 Feb 2026 21:04:32 +0000

i spend most of my time with claude thinking about when my daily usage limit is going to reset

New comment by czk in "I was banned from Claude for scaffolding a Claude.md file?"

czk — Fri, 23 Jan 2026 03:19:40 +0000

“After creating a new account, I can confirm the quota drains 2.5x–3x slower. So basically Max (5x) on an older accounts is almost like Pro on a new one in terms of quota. Pretty blatant rug pull tbh.”

lol

New comment by czk in "Tell HN: Claude session limits getting small"

czk — Thu, 22 Jan 2026 01:56:09 +0000

claude has half the context window size of codex and blows through a good percentage of it right off the bat by injecting a system prompt the size of don quixote

New comment by czk in "OpenAI API Logs: Unpatched data exfiltration"

czk — Wed, 21 Jan 2026 22:08:49 +0000

it sounds like the data can be involuntarily disclosed to an external third party (the attacker’s domain) purely because someone reviewed logs that auto-load remote images

their log viewer renders the markdown and their browser will make a request containing the sensitive data to the attackers domain where it can be logged and viewed