My work sends out phishing test emails, so I am reporting all work email as phishing, so I will get a perfect "score."
Craig Stuntz
I can write an incorrect program which is faster than your incorrect program
Astronauts: "I have two Microsoft Outlooks and neither one of those are working"
CAPCOM: "Switch to Mac, lol"
Astronauts: "Always has been"
On my walk this afternoon I spotted a Turkey Vulture, and as I recorded this observation in Merlin Bird ID I noticed that they are hiring a Python machine learning engineer https://jobs.hr.cornell.edu/us/en/job/WDR-00051454 Could be a fun job for the right person!
RE: @arstechnica@mastodon.social
This is particularly galling because iOS, unlike Android, has no notification history feature. So past notifications were available to the cops, but not to end users
I'm indebted to Aaron Stump for pointing out this paper on his podcast, which I'm enjoying quite a lot: "The Calculated Typer (Functional Pearl)"
https://al.radbox.org/doi/10.1145/3759164.3759346
"If you want a picture of the future of programming, imagine reinventing package managers, for ever."
-George Orwell (probably)
https://nesbitt.io/2026/03/08/if-it-quacks-like-a-package-manager.html
Amateur with ChatGPT successfully "vibe-maths" a 60 year old unsolved Erdős problem https://www.scientificamerican.com/article/amateur-armed-with-chatgpt-vibe-maths-a-60-year-old-problem/
"AI is often incorrect!" is not a useful framing. A useful framing is a confusion matrix, plus the ability to exclude false positives (e.g., with constraints). The same as any linter. Then you can judge the usefulness.
Nice video on making math in LaTEX documents converted to PDF accessible https://www.youtube.com/watch?v=1cRYxldSIh8
