As a lawyer would say "not suitable for the purpose sold".
For a story about how Oxide avoided them, see "Holistic boot", at https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Frfd.shared.oxide.compu...

["Really quite horrid" is British for "<expletive deleted/> piece of <expletive deleted/> junk"]

Librarians have a list of books in print: if someone asks for one that isn't in print, they can say "it's out of print if it exists", and/or "it hasn't been printed yet", ditto.. Followed by "did chatgpt tell you about it? If it did, you need to tell it to not to tell you fibs". (:-))

That can't happen

https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Ftailscale.com%2Fkb%2F1123%2F...

Having a couple of binaries in the git tree would be a kind act: I often do that with (much smaller!) project of mine to make it easy of people to try and use them. Of course, to create MRs, you'll still need to build and debug it yourself.

Let's make programming more fun by making it harder!

Yup, same as the feedback loops in "cold readings"

Charlie Stross(@cstross@wandering.shop) wrote, in Mastadon:
The LLMentalist effect: Large Language Models replicate the mechanisms used by (fake) psychics to gull their victims: https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fsoftwarecrisis.dev%2Flet...

The title of the paper is "The LLMentalist Effect: how chat-based Large Language Models replicate the mechanisms of a psychic’s con"

That's been done for years, just not by Facebook. It's a classic problem in machine learning, and they've either finally noticed ... or took a revenus hit! Not making money is an excellent motivator (:-))

Interests, like "I want a trailer hitch for my Subaru".

All that other stuff? Wanted for someone else.
It's just piggybacking the blame onto advertisers, because people don't like them anyway

• - they own the agent for the advertiser,
• - they own the agent for the publisher,
• - they own the auction house, and
• - they don't provide an audit trail.

I used to work in advertising, and I saw Google as the personification of "moral hazard" (which see). Other things? Way nicer.

If you scan a thousand British faces and compare them to a thousand criminals, you will do 1,000,000 comparisons. (that's the birthday paradox part).
If your error rate is 0.8%, you'll get roughly 8,000 false positives and negatives.
That's bad enough if they are all false positives: people get arrested, then released.
It's way worse if they are all false negatives: 8,000 criminals get ignored by the police dragnet.

That was Britain: false positives are life-threatening in countries where the police carry guns.
0.8% is a good error rate. 34% wrong is typical in matching black women. See
https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.aclu-mn.org%2Fen%2Fnews%2Fbiased-technology-automated-discrimination-facial-recognition%23%3A~%3Atext%3DStudies%2520show%2520that%2520facial%2520recognition%2520technology%2520is%2520biased.%2Cpublished%2520by%2520MIT%2520Media%2520Lab.

At a certain company long long ago, managers had a mainframe-based planning app that looked like a sort of spreadsheet.

The company did a study to see how much it improved the manager's team's productivity...

Oops! Use of the tools was correlated with declining productivity.

Alas, if a 16-year-old is doing the buying over the net, Bob's out of luck. Bob can't even tell if the customer is a dog (:-))
The parent who buys their kid a phone is a really good source of the kid's age. It's not leak-proof, but most other approaches leak like sieves.

Once upon a time, lint would critique your work, at the expense of often being wrong. I tried using LLMs for that, and they work only a little less well than lint.
I was insulted, though, when It misquoted my sentence and then said I had a grammar error in the misquoted part (:-))