Hacker News: LUmBULtERA

New comment by LUmBULtERA in "Claude is only available to people over 18 years"

LUmBULtERA — Fri, 11 Sep 2026 13:31:07 +0000

Yes, I'm wondering why it's upvoted like this is something new today. What's going on?

New comment by LUmBULtERA in "iPhone 18 Pro and iPhone 18 Pro Max"

LUmBULtERA — Wed, 09 Sep 2026 21:02:27 +0000

My 15 pro bought on release is at 91% and 607 cycles, for more reference.

New comment by LUmBULtERA in "Do you think it happened? Research stolen from their Codex private chats"

LUmBULtERA — Tue, 08 Sep 2026 20:08:19 +0000

Just an FYI, despite the wording on this, my "don't train on this" checkbox never changes (I don't need to re-toggle it, or change it on every device, etc...). This post reads as if its something that needs to be done repeatedly. Also if you turn on the advanced security option, ChatGPT locks to no training and cannot even be toggled on if you wanted it to.

New comment by LUmBULtERA in "OpenAI restores 5-hour Codex and Work limits for ChatGPT Plus users"

LUmBULtERA — Tue, 25 Aug 2026 14:05:23 +0000

How is that the case? ChatGPT not only gives a lot of Codex/Agentic usage, once you've run out of that you still have a massive quota of Chat usage that's entirely separate. You can still use plug-ins with Github, Google Drive, etc... using the Chat usage as well.

New comment by LUmBULtERA in "DeepSeek V4 Flash 0731"

LUmBULtERA — Fri, 07 Aug 2026 19:07:43 +0000

I'm still not sure, there's a promo going on now, but generally Go gives $60 of API credit and right now it might be $120 with deepseek. But $20/month OpenAI subscription I believe gives you many hundreds of API-equivalent usage? I've heard $100/month giving many thousands API-equivalent per month.

New comment by LUmBULtERA in "DeepSeek V4 Flash 0731"

LUmBULtERA — Fri, 07 Aug 2026 18:25:05 +0000

Is it still cheaper than Luna if using an OpenAI subscription? My gut is no, but I have not done the math.

New comment by LUmBULtERA in "Improving GPT‑5.6 Sol in ChatGPT, expanding GPT‑5.6 Luna access for free users"

LUmBULtERA — Fri, 07 Aug 2026 17:29:05 +0000

>They need to slow down and do a model release once a year and only do extremely minor iterations instead.

There are a ton of use-cases where the models still struggle a lot and make bad decisions, I'd much prefer they continue their acceleration for a while more.

New comment by LUmBULtERA in "Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber"

LUmBULtERA — Tue, 21 Jul 2026 18:40:08 +0000

I have friends and family who use Gemini, but entirely because their Pixel phones came with a year of it for free. No other reason, and they will most likely never pay for it.

New comment by LUmBULtERA in "Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber"

LUmBULtERA — Tue, 21 Jul 2026 18:37:43 +0000

The subsidizing thing is repeated over and over without proof. Personally, I doubt they're actually subsidized.

New comment by LUmBULtERA in "The Kimi K3 Moment"

LUmBULtERA — Sun, 19 Jul 2026 20:38:53 +0000

Maybe, but Tibo said something on Twitter last week making it sound like it might not be.

New comment by LUmBULtERA in "The Kimi K3 Moment"

LUmBULtERA — Sun, 19 Jul 2026 00:13:41 +0000

Given how OpenAI got rid of their 5-hour limits and reset weekly limits so often, is Kimi really undercutting them on effective price?

New comment by LUmBULtERA in "The Kimi K3 Moment"

LUmBULtERA — Sun, 19 Jul 2026 00:11:01 +0000

Sure, but at those API rates it doesn't beat the subscription plans for OpenAI.

New comment by LUmBULtERA in "ChatGPT Work"

LUmBULtERA — Thu, 09 Jul 2026 19:46:10 +0000

Literally no UI changes when I toggle between the two. It's 100% identical, one is not less technical than the other.

New comment by LUmBULtERA in "ChatGPT Work"

LUmBULtERA — Thu, 09 Jul 2026 19:39:38 +0000

>Also, when you toggle btween ChatGPT Work and ChatGPT Codex, nothing changes. This is super confusing.

This is the weirdest thing. It's completely unclear even if I was using the app for "work" rather than "coding" why I wouldn't just use the "Codex" option. Why even bother toggling? Can Codex actually not do any of the things in "Work"?

New comment by LUmBULtERA in "GPT-5.6"

LUmBULtERA — Thu, 09 Jul 2026 18:48:20 +0000

>They mean the same thing as GPT-5.6-Max, GPT-5.6-Plus, and GPT-5.6-Fast but require base knowledge for an analogy.

But do they though? When do you use GPT-5.6-Max-Low vs. GPT-5.6-Plus High? Or GPT-5.6-Fast-Xhigh? What's the Pareto optimal choice (outcome and price)? According to the benches it seems to bop around and the even if the benches are accurate the best choice isn't always consistent.

New comment by LUmBULtERA in "Ask HN: Fable 5 promo ends today – is anyone keeping it once it's usage-billed?"

LUmBULtERA — Tue, 07 Jul 2026 15:30:58 +0000

Ping ponging to ChatGPT/Codex for 5.6 launch to test that out... Hopefully its out today or soon.

New comment by LUmBULtERA in "GPT-5.6 Sol Ultra will be in Codex"

LUmBULtERA — Mon, 06 Jul 2026 10:03:02 +0000

There could be a whole host of reasons. It may be during launch that compute is re-allocated from training to inference so that all users can try Fable. Soon that compute will re-allocate back to training until they can get more compute.

New comment by LUmBULtERA in "Claude Sonnet 5"

LUmBULtERA — Wed, 01 Jul 2026 13:53:10 +0000

Huh, super interesting. Thanks!

New comment by LUmBULtERA in "Claude Sonnet 5"

LUmBULtERA — Wed, 01 Jul 2026 10:38:35 +0000

Did Anthropic have Opus 4.8 and Sonnet 5 switched in the Agentic Search chart at first?

New comment by LUmBULtERA in "Claude Sonnet 5"

LUmBULtERA — Tue, 30 Jun 2026 19:01:34 +0000

That's yet to be determined. I think a lot of open-weight models are benchmaxxed and their usefulness for many tasks are not represented by those.