the internal disaster recovery playbooks at major organizations are sprawling wikis written entirely by claude. when something is going horrendously wrong and employees are frantically reading through internal documentation to work out how to prevent total collapse, they will be squinting at instructions such as "The purpose of the narrower lever is what isolates it: harm flowing through the cross-level tier. What it gets you is what's saved: full user capture, where convergence is the target."
@jk As I read all these comments I feel like crying. I am recently retired; I wrote a lot of technical documentation in my day, and if I were told to let Claude et al do the writing ✍️ I would just burst into flames.
@jk
English translation:
One of the cross beams has gone out askew on the treadle.
@jk
Aww gad. Now you made me think about what business processes documentation must look like nowadays. Pity the ISO 9001 auditor.
@jk Yes, these tools magnify, not replace, the efforts of writers and coders. Many do not understand this.
@jk “The smoking gun? There the entire time, completely unnoticed. The real issue is a one line fix.” and then it moves to the next troubleshooting section
@jk just out of curiosity, how much output have you been exposed to?
@jk this is so accurate it’s upsetting
@jk qwen inherits the same cryptic speak
@lritter whether they say so officially or not, i'm certain its heavily distilled from claude. it could be a kind of emergent thing that happens when RLing models for coding performance though, by encouraging broader sets of semantic encodings that allow a greater variety in conceptual leaps via synonyms and odd phrasing, at the expense of producing output that humans find increasingly disgusting. but my money's on it just being a distillation thing
@jk they use training material sampled from claude, this is commonly known, and is also evident from just asking qwen about itself or its "co-authored by claude" commit messages. although anthropic now hides reasoning traces everywhere, so can't do that anymore.
imo this strangely compressed english is an artifacts of models optimizing inter-context talk (bot 2 bot). b/c if you ask it later to explain in comprehensible english, it can do so. (cont)
@jk why this is happening i think is because they write under the assumption that humans aren't going to read docs or comments anyways. so if you want a different style, you have to explicitly say so, or fix it in post.
@jk when I did datacenter work, I printed and tested the DR plans. Now with the clouds, there’s less for me to do other than wait and check status. I still test the plans.I don’t print because if I can’t login to view the DR plan, I have to wait to use it anyway (and it’s so short I have it memorized).
Testing DR plans is part of several common certifications. I don’t doubt there are bad ones, but it likely won’t be a surprise. It’s valid to test a DR plan and find it to be bad.