[FRIAM] replication crisis?

[glen]

Last insomnia-driven nonsense post ... I promise. From this paywalled article:

https://www.newscientist.com/article/2584476-quantum-computing-may-be-facing-a-replication-crisis/

to this paper:

Works on My QPU: Reproducibility in Quantum Computing Research
https://arxiv.org/abs/2607.08348

My own workflows are stupid simple. But even there, I struggle to balance simplicity and efficacy so that others might be able to execute them (were anyone ever interested in doing so, which they're not). And the AI harnesses make it worse. You see it in Marcus' last report: "Paths are relative to this document’s directory" Etc. In my last commit yesterday, the agent put the *symlinked* structure into the .gitignore file. Yes, symlinked. So while it wrote scripts that assumed a directory structure, because they were symlinked (and we discussed why - the bastard knew full well that one of the code bases is horizontal across projects) put them in the .gitignore. Sorry. Obviously, it couldn't know how much or why I depend on `git status`.

I desperately want to use something like this https://wholetale.org/#featured. But it's an overhead that's often out of scope ... even if they're comfortable with FAIR+, about which we see more lip service than sincere commitment.

-- 
8647 ⊥ ɐןןǝdoɹ ǝ uǝןƃ
ὅτε oi μὲν ἄλλοι κύνες τοὺς ἐχϑροὺς δάκνουσιν, ἐγὰ δὲ τοὺς φίλους, ἵνα σώσω.