Tootfinder

from my link log —
Don't make my mistakes: common infrastructure errors i've made.
https://matduggan.com/mistakes/
saved 2021-12-03 https://dota…

As I edit my writing, I've found it useful to have different layouts to catch mistakes. Earlier this year I moved to Obsidian for my a writing platform since I can tweak it with Claude or Codex. Here's an example of such a layout I created. It's in HTML, so I can't edit it (... for now).

OMG, I'm having a laugh with Opus 4.8. They trained it to be more honest, but now it's like Jim Carrey in Liar Liar. Earlier it was like "I accidentally deleted all the work I just did, but it's okay - I fixed it." And now it's admitting every little mistake it makes. It's hilarious.

I remember watching the 2022 documentary
"Aşk, Mark ve Ölüm"
(Love, Deutschmarks and Death) by Cem Kaya
with awe and shame at how much I had not known.
The film looks at the difficult lives of Turkish workers in Germany through the music they produced.
One widely popular song, sung by Yüksel Özkasap, included the lines:
“Germany, everything about you is a lie. You have stolen my life and taken my love.”
The language textbook
"Türk…

Yeesh. I tried to do an article merge on the Spanish #Wikipedia because it somehow ended up with duplicate articles on the same subject.
Obviously this is pretty hard to do because I barely understand #Spanish and I did the mistake of turning one page into a redirect which is apparently a no-no judg…

Worked on a small hobby site yesterday that I had not really touched since refactoring it onto #11ty in 2023 and realised I had made a ton of mistakes and had not finished parts of it off! Still not finished but built better using 3 .json files with data in instead of 60 .md files. And also stopped the build function on Netlify - just deploys the files now and in only 6 seconds from a sync to G…

I just got the mail about this and what the actual fuck?
https://bsky.app/profile/did:plc:f2uxgfbnehk4e7bymnpodyn7/post/3mmujqtjmoc2j

“A lot of people are under the mistaken idea that the only way that you get productivity from A.I. is by removing labor costs.”
https://www.nytimes.com/2026/05/29/business/economy/ai-jobs-productivity.html?unlocked_article_code=1.mFA.3d4j.IAUrghRcyvk-&smid=nytcore-ios-share

How does an LLM know the difference between its own thoughts
and someone else's words?
https://role-confusion.github.io/

Prompt Injection as Role Confusion
LLMs can't tell who's speaking. We show they identify roles by writing style, not tags, and exploit this with CoT Forgery, injecting fake reasoning that models mistake for their own thoughts.

Crosslisted article(s) found for cs.OS. https://arxiv.org/list/cs.OS/new
[1/1]:
- A Secure, Manifest-Based Framework for Delegated Privilege Promotion
Rajarshi Chowdhury, Akshay Shah
https://arxiv.org/abs/2605.28991 https://mastoxiv.page/@arXiv_csCR_bot/116656780860008224
- IORM: Hierarchical I/O Governance for Thousands of Consolidated Databases on Oracle Exadata
Rajarshi Chowdhury, Akshay Shah, Zakaria Alrmaih, Chenhao Guo, Anubhav Singh, Sue Lee
https://arxiv.org/abs/2605.29006 https://mastoxiv.page/@arXiv_csDB_bot/116656694349323904
toXiv_bot_toot