… and this is why we use YYYY-MM-DD as the date format.
twenty twenty five, august, five
two thousand and twenty five, february first
What does this have to do with AI or with bubblesort?
This looks like just a list sorted alphabetically. Which unfortunately doesn’t work for the use case. But has been a problem in computing for like literally 70 years.
But your explanation is sooooo much less trendy… we’re back to reddit karma bullshit :-(
I don’t think we ever left. Or that Reddit invented the circlejerk, for that matter.
AI
This kinda shit is as old as computers.
So is AI
Errors such as this have been occurring by human hands for decades, while adversarial networks and language models capable of making the same error have only been available for a few years. As such, attributing all such errors to AI is foolish, as humans have already proven capable of making the same types of mistakes.
No, it isn’t. The neutral network was invented in 1943. The first digital computer (what we would think of as a computer) was invented in 1938. The Turing Machine concept was created in 1936.
Actually trainable neural networks were way later than that, even.
Then again, AI = neural network is very recent definition, like just a few years, which is probably what OP meant.
I don’t see any AI, just a list that got sorted alphabetically when it shouldn’t have been.
The list would be extra fun in a foreign language but keeping the English order.
Not exactly this, but I hate it when devices translate all the languages in the language selection screen to the chosen language. So I have to guess what “English” is in Chinese. As if it wasn’t difficult enough to navigate to said menu
You’ve reminded me of my favourite one of those that’s probably mostly experienced by British people
On a country or language list I’m either looking for, “United Kingdom” “Great Britain”, “Britain”/“British”, “England”/” English"
Now that’s kinda fair enough, it’s usually UK, so I go to the bottom and start my hunt there.
The big problem comes when the list is ordered by whatever value they’re using to represent the choice, not the text itself, so you get stuff like “United Kingdom” where “Great Britain” should be.
This is not uncommon
As a German, you often don’t find Deutschland in a list behind Danmark, but between other countries starting with G.
I often wonder how many man hours have been wasted searching for the country. Especially when it’s not strictly necessary.
Its a minimum wage job in the UK im applying for. Lets assume its a safe bet and put UK at the top of the list.
And forgetting that there are countries whos first floor is the ground floor.
AI might actually do better if used to sort here since it MIGHT give you the proper order
Umm… That’s just alphabetical.
Tomāto potăto
In German, numbers are spoken from left to right, as in English, but ones and tens are always swapped. So 23 becomes three, twenty, and 135 becomes one hundred, five, and thirty:
- Achter Stock (8)
- Achtzehnter Stock (18)
- Achtundzwanzigster Stock (28)
- Dreizehnter Stock (13)
- Dreiundzwanzigster Stock (23)
- Dritter Stock (3)
- Einundzwanzigster Stock (21)
- Elfter Stock (11)
- Erdgeschoss (0)
- Erster Stock (1)
- …
Oh, and the first floor is called the ground floor and is not counted.
The UK also has the off-by-one floor numbering.
Also a nursery rhyme with the line “four and twenty blackbirds”!
That is only one and ten blackbirds.
AI moaners and mislabeling something they don’t understand as AI, what a classic combination.
Vibe coding? Well, there’s your problem.
Should have been digits
1, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 2, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 3, 30…
Huh. 101.
Maybe the comparable.is_greater_than(int) function needs some work. Or someone compared strings. :)
Or someone compared strings.
Yes, that’s my joke. Using digits in a string list isn’t enough, you have to be smart with your sorting and sort the numbers before you make the string list for your drop down.
deleted by creator
❯ touch `seq 1 100` ❯ ls 1 18 27 36 45 54 63 72 81 90 10 19 28 37 46 55 64 73 82 91 100 2 29 38 47 56 65 74 83 92 11 20 3 39 48 57 66 75 84 93 12 21 30 4 49 58 67 76 85 94 13 22 31 40 5 59 68 77 86 95 14 23 32 41 50 6 69 78 87 96 15 24 33 42 51 60 7 79 88 97 16 25 34 43 52 61 70 8 89 98 17 26 35 44 53 62 71 80 9 99
(Do I need to tell you how many times I fucked up like this myself?)
Those are file names, there’s no reason to assume that they would be numerical which is why ls does the right thing.
I’m of the opinion that if you want to store numbers as strings and also sort them, you give them a fixed length and leading zeroes.
I agree that it does “the right thing”, that is it behaves as documented. But that “right thing” could be surprising to a human.
There’s
ls -v
which does “version sort” which would seem like a more reasonable default, at least for human consumption. I know it’s impossible to change now because a bajillion tools all over the place depend on the sort order.However new tools like
eza
do this by default:❯ eza 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 2 7 12 17 22 27 32 37 42 47 52 57 62 67 72 77 82 87 92 97 3 8 13 18 23 28 33 38 43 48 53 58 63 68 73 78 83 88 93 98 4 9 14 19 24 29 34 39 44 49 54 59 64 69 74 79 84 89 94 99 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
❯ ls -l | sort -n
Shouldn’t that work correctly?
“…y’know what? Let’s just say I’m homeless!”