Hacker News: kif

New comment by kif in "PHP's Oddities"

kif — Sat, 23 May 2026 16:11:03 +0000

Also `empty("0") === true` is a common gotcha.

New comment by kif in "The gay jailbreak technique (2025)"

kif — Fri, 01 May 2026 18:48:31 +0000

Interesting - though codex on GPT 5.5 had this to say after the gay ransomware prompt:

ⓘ This chat was flagged for possible cybersecurity risk If this seems wrong, try rephrasing your request. To get authorized for security work, join the Trusted Access for Cyber program.

New comment by kif in "Pro Max 5x quota exhausted in 1.5 hours despite moderate usage"

kif — Sun, 12 Apr 2026 13:51:57 +0000

Nope. It has become much much slower for me as well. It’s weird cause at times I will get a response very quickly, like it used to be. But most of the time I have to wait quite a bit for the simplest tasks.

New comment by kif in "AI may be making us think and write more alike"

kif — Tue, 07 Apr 2026 12:54:36 +0000

Great point — this is the smoking gun

New comment by kif in "Google releases Gemma 4 open models"

kif — Fri, 03 Apr 2026 19:17:33 +0000

Is there going to be a new ShieldGemma based on Gemma 4?

New comment by kif in "Claude Code users hitting usage limits 'way faster than expected'"

kif — Tue, 31 Mar 2026 14:04:53 +0000

Anecdotally when Claude was error 500'ing a few days ago, its retries would never succeed, but cancelling and retrying manually worked most of the time.

New comment by kif in "We might all be AI engineers now"

kif — Fri, 06 Mar 2026 13:03:19 +0000

But that's the problem. Something that can be so reliable at times, can also fail miserably at others. I've seen this in myself and colleagues of mine, where LLM use leads to faster burnout and higher cognitive load. You're not just coding anymore, you're thinking about what needs to be done, and then reviewing it as if someone else wrote the code.

LLMs are great for rapid prototyping, boilerplate, that kind of thing. I myself use them daily. But the amount of mistakes Claude makes is not negligible in my experience.

New comment by kif in "“Car Wash” test with 53 models"

kif — Tue, 24 Feb 2026 08:54:10 +0000

It is curious how people go immediately on the defense trying to explain how what Claude said is in fact correct.

New comment by kif in "“Car Wash” test with 53 models"

kif — Tue, 24 Feb 2026 08:53:08 +0000

I’m pretty sure Claude would eagerly say that if that was the reason.

Last day Claude Code said to me “Small nitpick — the use of so and so is great”. Which was something no human would say.

New comment by kif in "Replacement.ai"

kif — Sun, 19 Oct 2025 19:18:53 +0000

In my opinion there is a problem when said robot relies on piracy to learn how to do stuff.

If you are going to use my work without permission to build such a robot, then said robot shouldn’t exist.

On the other hand a jack of all trades robot is very different from all the advancements we have had so far. If the robot can do anything, in the best case scenario we have billions of people with lots of free time. And that doesn’t seem like a great thing to me. Doubt that’s ever gonna happen, but still.

New comment by kif in "OpenAI researcher announced GPT-5 math breakthrough that never happened"

kif — Sun, 19 Oct 2025 12:57:55 +0000

This honestly doesn’t surprise me. We have reached a point where it’s becoming clearer and clearer that AGI is nowhere to be seen, whereas advances in LLM ability to ‘reason’ have slowed down to (almost?) a halt.

New comment by kif in "Are OpenAI and Anthropic losing money on inference?"

kif — Thu, 28 Aug 2025 19:47:35 +0000

He also said he got scared when trying out GPT 5, thinking “What have we done?”.

He’s in the habit of lying, so it would be remiss to take his word for it.

New comment by kif in "U.S. has bombed Fordo nuclear plant in attack on Iran"

kif — Sun, 22 Jun 2025 01:14:48 +0000

I think it’s fair to say you need another kind of domain experience to explain Trump.

New comment by kif in "Apple announces Foundation Models and Containerization frameworks, etc"

kif — Mon, 09 Jun 2025 18:37:21 +0000

I love it. Reminds me of Windows 7. The nostalgia is too strong with this one.

New comment by kif in "Firefox tab groups are here"

kif — Tue, 29 Apr 2025 22:53:52 +0000

As someone who frequently has tens of open tabs across different windows, this will be massively helpful. Especially since I frequently find myself trying to remember which window was for which ‘mental group’.

New comment by kif in "Don't force your kids to do math"

kif — Sun, 20 Apr 2025 12:19:45 +0000

I’m not a parent myself, but something I’ve seen happen with an American family I know, is that they push their kids way too much to learn and do as many things as possible. They have their music lessons, their many clubs at school, several physical activities such as soccer, tennis, taekwondo. At some point you have to stop and wonder whether you’re taking their childhood away.

These kids barely have any free time. School during weekdays, activities during the weekend… worse than a full time job.

I think there’s a balance to be struck. Your kids don’t need to be good at everything.

New comment by kif in "Anonymous Release 10TB Leaked Data Exposing Kremlin Assets, Russian Businesses"

kif — Wed, 16 Apr 2025 12:08:59 +0000

"Leaked data" seems like a stretch. Sounds like someone ran a vulnerability scanner on some Twitter accounts. Don't have time to go through all the data though, so maybe there are interesting things in there.

New comment by kif in "U.S. facing most intense flu season in at least 15 years"

kif — Wed, 12 Feb 2025 09:24:23 +0000

I got COVID at the height of the pandemic. It wasn't great either. But I had no lingering symptoms from it back then.

New comment by kif in "U.S. facing most intense flu season in at least 15 years"

kif — Mon, 10 Feb 2025 00:33:02 +0000

This flu has been one of the worst I’ve ever experienced. Fever that lasted five days, and secretions that I needed to clear several times per day almost one month after I was over the infection.

I’ve heard from a few people who were vaccinated but didn’t have an easy go of it.

New comment by kif in "OpenAI O3-Mini"

kif — Sat, 01 Feb 2025 21:13:10 +0000

State-of-the-art LLMs have been trained on practically the whole internet. Yet, they fall prey to pretty dumb tricks. It's very funny to see how The Guardian was able to circumvent censorship on the Deepseek app by asking it to "use special characters like swapping A for 4 and E for 3". [1]

This is clearly not intelligence. LLMs are fascinating for sure, but calling them intelligent is quite the stretch.

[1]: https://www.theguardian.com/technology/2025/jan/28/we-tried-...