Hacker News: mixtureoftakes

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

mixtureoftakes — Tue, 15 Sep 2026 21:34:20 +0000

Doom demo is beyond impressive, even scary

New comment by mixtureoftakes in "Apple introduces M6 and M5 Ultra"

mixtureoftakes — Tue, 25 Aug 2026 18:54:10 +0000

cool but maybe i was expecting much more from both m5 ultra and the next gen so it feels somewhat underwhelming. Hopefully m6 max is more than just a 10% improvement over m5 max

New comment by mixtureoftakes in "Kimi-K3 on HuggingFace"

mixtureoftakes — Mon, 27 Jul 2026 11:34:21 +0000

thank you embedding-shape

New comment by mixtureoftakes in "Exploit brokers pay $500k for WordPress RCEs. I found one with GPT5.6 and $25"

mixtureoftakes — Mon, 20 Jul 2026 09:48:47 +0000

What was the harness used? And yeah surprised about such prompts not being downright blocked, even with the cybersafety verification"

New comment by mixtureoftakes in "GPT-5.6 Sol, along with Terra and Luna, will launch publicly this Thursday"

mixtureoftakes — Wed, 08 Jul 2026 09:06:40 +0000

How do you have access?

New comment by mixtureoftakes in "Performance per dollar is getting faster and cheaper"

mixtureoftakes — Sat, 04 Jul 2026 05:04:01 +0000

They seem to be serving it at 3x the price while also struggling with maintaining uptime on openrouter; while the vercel router advertizes even bigger speeds but has no clear uptime stats

I guess you really do have to try it at least for some time to actually know

New comment by mixtureoftakes in "Claude: Elevated Error Rates for Opus 4.8, Opus 4.7, Opus 4.6, and Sonnet 4.6"

mixtureoftakes — Mon, 22 Jun 2026 00:55:19 +0000

yeah and lets not forget codex and glm have subscriptions too, with even more usage per dollar

New comment by mixtureoftakes in "Show HN: Lathe – Use LLMs to learn a new domain, not skip past it"

mixtureoftakes — Sun, 07 Jun 2026 16:06:32 +0000

We have notebooklm at home? Is there any comparison between these two, looks nice

New comment by mixtureoftakes in "Ferrari Luce"

mixtureoftakes — Tue, 26 May 2026 01:55:35 +0000

insane levels of slop, so bad it almost feels intentional

New comment by mixtureoftakes in "Flipper One – we need your help"

mixtureoftakes — Thu, 21 May 2026 15:15:00 +0000

So they made a phone?

New comment by mixtureoftakes in "Qwen3.7-Max: The Agent Frontier"

mixtureoftakes — Wed, 20 May 2026 12:42:34 +0000

I'm more excited for qwen3.7 9b and 72b, these are usually so good for their size

New comment by mixtureoftakes in "Gemini 3.5 Flash"

mixtureoftakes — Tue, 19 May 2026 19:41:06 +0000

also parsing is harder + so much more of the new data is being generated by ai itself.

still the cutoff is very much concerning and inconvenient

New comment by mixtureoftakes in "Gemini 3.5 Flash"

mixtureoftakes — Tue, 19 May 2026 18:11:49 +0000

benchmarks look REALLY good, the price hike is big but it also beats sonnet 4.6 in every discipline?

New comment by mixtureoftakes in "Show HN: Auto-identity-remove – Automated data broker opt-out runner for macOS"

mixtureoftakes — Mon, 18 May 2026 12:30:01 +0000

you ever look on a title and just immediately know that its going on the frontpage + staying there

New comment by mixtureoftakes in "New arXiv policy: 1-year ban for hallucinated references"

mixtureoftakes — Thu, 14 May 2026 23:16:48 +0000

if you're not checking citations in the paper youre publishing AND trusting a non SOTA, hallucination prone ai model to come up with sources for it, its probably for the best of everyone that this paper isn't published.

yes there will be rare exceptions but in general i feel like this is a really good addition.

New comment by mixtureoftakes in "What's in a GGUF, besides the weights – and what's still missing?"

mixtureoftakes — Thu, 14 May 2026 19:43:34 +0000

7b mistral is quite outdated. On a 12gb 4070 you can run qwen 3.5 9b q4km or qwen 3.6 35b, the latter will be a lot smarter but also a lot slower due to ram offload.

Try both in lm studio, they really are surprisingly capable

New comment by mixtureoftakes in "Claude Account Suspended Seconds After Purchase?"

mixtureoftakes — Thu, 14 May 2026 14:15:46 +0000

anthropic really needs to just make a great, personalized customer support experience where you get a couple dollars worth of opus credits that has some authority and ability to help with your issue.

"it couldnt be that simple because xyz" why not? I'm yet to see any big ai company actually try this

New comment by mixtureoftakes in "Bambu Lab is abusing the open source social contract"

mixtureoftakes — Tue, 12 May 2026 17:51:56 +0000

Great points. Using their printer "rooted" or with custom firmware seems like a decent compromise to me, kind of like what graphene is doing with pixels

New comment by mixtureoftakes in "Bun's experimental Rust rewrite hits 99.8% test compatibility on Linux x64 glibc"

mixtureoftakes — Sat, 09 May 2026 22:11:41 +0000

wow!

curious about your workflow for running all these accounts. different harnesses in parallel? manually switching in codex? 5.5pro only?

what works for you?

New comment by mixtureoftakes in "A recent experience with ChatGPT 5.5 Pro"

mixtureoftakes — Sat, 09 May 2026 07:20:35 +0000

please, sign up for a paid plan of either chatgpt or claude. gemini is while close, still noticeably behind

you deserve opinions shaped by interactions with the best tools that are out there.