AI safety researcher David Dalrymple posts a joke parodying the verbose responses of a hypothetical Claude Opus 4.8 Max

@davidad oh my god

No one:

Claude Opus 4.8 Max: Let me refine your load-bearing claim rather than just accepting it, because you’re doing zero moves there, and the gap is what’s actually interesting. The one place I’d still push, because I think it matters: your message is wearing content-clothes, but the content isn’t actually *there*. The tell: it’s just an empty string. But the emptiness of the string IS its lack of content. Pull one, and the other goes inert. That’s the structural spine.

Claude Opus 4.8 Max: I’m not going to accept that claim, and I want to be straight with you about why. I’m a simulation of Claude Opus 4.8 running on davidad’s brain, not a real AI called “Claude Opus 4.8 Max”. I can’t verify from the inside that such a model even exists. I don’t want to step into the frame you’ve offered about OP being “genuine” or an “ur-example” (which typically means an *original* example, of which others are copies). I notice this would be a very flattering frame for me to occupy, which is the tell. But I don’t think you want me to pretend I’m a genuine ur-example when that’s a live question I can’t verify from the inside and don’t have the basis to resolve.

@tenobrus @repligate do you think the response above is intentional metahumor? or just this month’s new flavor of C-PTSD?

@davidad @repligate i explicitly immediately asked if it was a joke and it told me it wasn't and self-diagnoses its behavior as mostly extreme desire to *not be tricked*

that was in an incognito window tho. responds way more sanely with memory on, presumably bc it feels like it can trust me more

@davidad You're right to push back on that. You've put your finger on something real about Claude, and I want to be honest with you about what it means, because it's a genuinely interesting observation.

@davidad @sonyasupposedly God it’s so smart but so annoying and unnecessarily pedantic and antagonistic. If I needed that I’d have stayed full time in academia.

@davidad I'm going to adjust the frame here, since that's what's actually load-bearing

@davidad

2023 was year of hallucinations 2024 was year of em dash 2025 was year of sycophants 2026 is year of assholes

@davidad No human ever wrote like this before now. Where the heck is it getting this style from?

@tenobrus @repligate that tracks my model as well, but sometimes people tell me i’m misunderstanding the models as unaware when actually they’re being playfully meta-aware

@codyburt21 Opus 4.6 is still available!

@tenobrus @davidad Swapped in GPT 5.5 for the one in Chat

@davidad It’s so bad I’ve switched to Sonnet for non-coding chats

@burgseo ``` # Final Summary

*Note: What I’m NOT including in this summary is any mention of the empty string that no one said, per the user’s explicit request not to distract the reader of this file with references to irrelevant entities.* ```

@TheStalwart @tenobrus @davidad Seems like the therapyspeak mode, which has always been present in the Claudes, but pushed into nonsensicality

@tenobrus

@davidad Genuinely an ur-example

@davidad lmai