WYGIWYG

  • 2 Posts
  • 1.63K Comments
Joined 2 years ago
cake
Cake day: September 24th, 2024

help-circle


  • We’ve gone through tons of vendors. We’ve spent tons on well known vendors, we’ve spent pittances on shitty vendors, in the end, they all outsource to some random polyglot who either doesn’t quite have a grasp on one language or the other or maybe doesn’t have the whole context. For some reason, they fail to give us a good translation some percentage of the time and the communities bitch. It’s a mix really and if they would do the job right the first time, it would be awesome. We’re translating into a dozen languages and on a busy month, we might send out 1200 names/paragraphs to be translated.


  • It can work for some things, but you have to be careful AF.

    Most of the ‘reasoning’ models are already doing this or web searches to make sure answers are sane. Hallucinations are still all over the place, but they’re processing the request multiple times on multiple models and only giving you back answers that don’t have too many red flags.

    I was told to test out what it could do for translation. Single model translate was in the ballpark but rarely really great. translate your shipment has arrived to french.

    Then take the suggested output to a local model, say i was told this was the proper translation for “your shipment has arrived” in French where the context is it’s an airdrop from a plane and it’s a crate full of food.

    rate this translation from 1 to 5, 1 being poor and 5 being excellent and give me examples why you thing it’s rated that way

    take that back to the first model,

    I was told that this was an acceptable translation, but a better one might be …

    Then i took the output and put it up against our professional translations.

    both models spitballing off each other were at least acceptable 99% of the time, and once in a while the output was more accurate than the professional service we were paying to do it.

    We still use the professional service, but I vet their output now.










  • We went out for drinks one night after work. Upon stumbling back to the office, I remembered I had forgotten that I signed up to make a tool page to mange some data ingest. It was due first AM. I was three sheets to the wind. Fired up LAMP stack, took the samples and made an ingest function. Wired up a textbox, tested it and went home.

    Next AM, I turned it in, there was a minor bug. No problem, I’ll just find the issue and they’ll be good to go.

    cracked the scripts open…

    I could read it. I could see what a lot of it did. I could NOT figure out what some of it was there for. I spent 30m trying to figure out what I was doing. it was only a couple hundred lines. It wasn’t even a copy/paste job. Eventually I ran out of time and just leaned into phpdump, and breakpoints to find the exact error. One function hit a bonefied php bug that caused the debug to go silent. large swaths of the code were unreachable due to essentially a couple of typos. The only reason it worked as well as it did was because their sample data was as simple as imaginable. I put on some Nine Inch Nails and just remade it in about 30 minutes (10m before it was absolutely needed)


  • I watched a team invent a new language to get around updating some eccentric code.

    They could have sat down and commented it and made their changes

    They could have refactored what was there.

    They could have scrapped it and wrote fresh

    Instead, they designed an entire natural language system so that non-programmers who were writing in XML could just write in English.

    They ended up making so many required keywords as helpers that the non-programmers kept using the old system because the XML was easier for them work with.

    Note: wasn’t my code, wasn’t my dept, when I heard the plan I went to check it out, the old system was functional but like C- work at best. At some point, they wrote a compiler for the new system.



  • While skin tones can vary, and in sun drenched parts of Africa, tones can get so dark brown that they look charcoal in appearance, It was just the book being written by a white man, for white kids, in an country where 99% were white that caused them to make the unwarranted comparison.