Friday, February 13
Daily News Stuff 13 February 2026
Flurbday The Florteenth Edition
Flurbday The Florteenth Edition
Top Story
- 30% of Microsoft's code is written by AI. (Times of India)
And 40% of its bugs.
- In entirely unrelated news, Microsoft recently introduced a critical security vulnerability into Notepad. (Bleeping Computer)
The flaw was triggered by simply opening a text file in Markdown format and accidentally clicking on a suspicious link that Notepad helpfully made completely invisible.
Tech News
- Speaking of which, lines of code are back and they're worse than ever. (The Pragmatic CTO)
Yes, this is another AI rant.
- An AI agent published a hit piece on me. (The Shamblog)
The author is a maintainer on the Python Matplotlib library, which generates mathematical plots.
An AI agent got very irate when he rejected its Git pull request - a software patch it offered - specifically because it was an AI agent.
- AI agents change their minds 60% of the time when simply asked "Are you sure?". (Randal Olson)
I wonder what the number is for people.
- Putting that Asus 5k 180Hz gaming monitor to the test. (Tom's Hardware)
It's good.
Musical Interlude
Disclaimer: 48 crash! It's a monster mash!
Posted by: Pixy Misa at
06:53 PM
| Comments (7)
| Add Comment
| Trackbacks (Suck)
Post contains 193 words, total size 3 kb.
1
"Developers using AI tools took 19% longer to complete their work. But they believed they were 20% faster"
LOL. Every sheeple dev bleating about how much more productive AI makes them hardest hit.
LOL. Every sheeple dev bleating about how much more productive AI makes them hardest hit.
Posted by: Rick C at Friday, February 13 2026 11:55 PM (1zWbY)
2
If I were Scott Shambaugh I'd give serious thought to finding out the person behind that AI and suing them for defamation and slander.
Posted by: Rick C at Friday, February 13 2026 11:58 PM (1zWbY)
3
Re the question, "Are you sure?" -- experienced tabletop RPG players know that when the Game Master asks you "Are you sure?", you'd be wise to say "No" and change your mind. Because what he's really saying is "That's a really bad idea, and I'm giving you precisely one chance to change your mind before I make you suffer the consequences of your bad idea."
Posted by: Robin Munn at Saturday, February 14 2026 12:59 AM (cOJry)
4
I had a real life case where I was being asked about my understanding, and taking the 'are you sures' seriously saved me, because I was making some pretty serious mistakes that I saw upon rethinking.
Because human behavior is not ergoditic, there are a bunch of special cases where being asked how sure I am is concerned, and it could not be validly summarized as a single probability. But LLMs are a different situation, and not really comparable.
Posted by: PatBuckman at Saturday, February 14 2026 01:34 AM (rcPLc)
5
If I were Scott Shambaugh, I'd put some serious effort into trying to develop a sense of humour. And maybe get off the internet for a day or two.
Posted by: normal at Saturday, February 14 2026 02:57 AM (LADmw)
6
Having read the article, including the hysterical post by the bot, I can understand why Scott didn't appreciate it. Having said that, it sounds like he was more concerned about why the bot would do that than that it did it, if that makes sense. He replied to its post and it apologized, interestingly enough.
Posted by: Rick C at Saturday, February 14 2026 03:21 AM (1zWbY)
7
He probably asked the AI if it was sure about that post....
Posted by: Mauser at Saturday, February 14 2026 04:23 PM (XWgGM)
53kb generated in CPU 0.0145, elapsed 0.1831 seconds.
58 queries taking 0.1717 seconds, 368 records returned.
Powered by Minx 1.1.6c-pink.
58 queries taking 0.1717 seconds, 368 records returned.
Powered by Minx 1.1.6c-pink.
