News for hackers. Stuff that matters.
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by quotemstr on Thursday September 03, 2026 @01:01PM (#49553183)
It's a good metaphor and I refuse to let AI ruin it for me.
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by spudlyo on Thursday September 03, 2026 @01:55PM (#49553957)
Years ago, when I worked at Stripe (which had a somewhat unique and inventive lexicon) it was a common term. “Is this jank load-bearing?” someone might ask.
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by cobzilla on Thursday September 03, 2026 @01:47PM (#49553860)
I added a specific rule to disallow saying “load bearing”. So Claude is now saying “load handling”
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by jazzyjackson on Thursday September 03, 2026 @06:18PM (#49557906)
Be sure to avoid asking it to ignore the elephant
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by andrewla on Thursday September 03, 2026 @02:17PM (#49554259)
Kids In The Hall had a sketch about overuse of a word or phrase [1]. This is the world that Claude is building for us.

[1] https://www.youtube.com/watch?v=lStcwT_RGrQ

Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by wadayano on Thursday September 03, 2026 @01:07PM (#49553281)
But have we found the seams yet though?
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by rl3 on Thursday September 03, 2026 @01:31PM (#49553630)
If the seams bear too much load, they rip. Whereas pants, they fall down.

I'll be honest with you: this is why we need to take a belt-and-suspenders approach.

Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by pborenstein on Thursday September 03, 2026 @01:52PM (#49553930)
That's not just an observation, it's an insight. Words are doing the real work.
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by HarHarVeryFunny on Thursday September 03, 2026 @02:50PM (#49554821)
This just makes me angry!

I really wonder if they can fix it. Fable 5.1 claims to speak humanese, but we'll see.

I tend to think this wierd limited vocabulary/style they use is an unwanted side effect of all the the RL training, perhaps also of being trained on their own synthetic content over multiple training cycles.

Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by wadayano on Wednesday September 16, 2026 @03:49AM (#49723286)
It’s not just limited vocabulary, it’s low key gaslighting. And honestly you’re so real for that!
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by cootsnuck on Thursday September 03, 2026 @01:36PM (#49553693)
Yea I don't get "load bearing" that much but "seams"... So sick of it.
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by SkyeCA on Thursday September 03, 2026 @02:39PM (#49554614)
It doesn't have to ruin it for you, but people are going to assume comments with it are AI generated.
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by necovek on Thursday September 03, 2026 @03:04PM (#49555043)
I can you can always use an em-dash instead of the hyphen for extra LLM cred: "load—bearing" :)
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by lubujackson on Thursday September 03, 2026 @01:19PM (#49553450)
We need a Samwise "Bear the load!" meme
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by BrokenCogs on Thursday September 03, 2026 @01:33PM (#49553660)
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by hirako2000 on Thursday September 03, 2026 @01:45PM (#49553828)
Content not viewable in your region.
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by BrokenCogs on Thursday September 03, 2026 @02:35PM (#49554548)
:(
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by MadameMinty on Thursday September 03, 2026 @01:35PM (#49553680)
Wait, why would it ruin it?
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by ipsod on Thursday September 03, 2026 @01:40PM (#49553761)
Because it says it more often than kids say "six seven".
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by tatersolid on Saturday September 05, 2026 @07:45AM (#49575629)
My kids stopped saying “six seven” about a year ago. I still have no clue what that was about.
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by ipsod on Sunday September 06, 2026 @02:26PM (#49589366)
I think it might be like, a praise of inanity.
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by imwally on Thursday September 03, 2026 @01:40PM (#49553756)
It’s a frequently used metaphor in LLM responses.
Re:Ask HN: Why are OpenAI, Claude, and Grok simultaneously d... by bornfreddy on Thursday September 03, 2026 @02:22PM (#49554334)
Often for trivial things that LLM is proud that it has noticed but bear no load whatsoever.