alt Hacker NewsWhen I tried to use CoPilot at my day job, I set up a prompt to process scans of checks and invoices (with the MICR account bar code line covered w/ a pen so that there was no PII), it would sort of work for 20 files at a time to write a .bat file command to rename the source image based on the content of the check and invoice --- this removed the need to open up scans in batches of 50, read the Invoice ID, save using that, then run a .bat file to rename them based on the balance of information in the database.
The thing is, the failure rate was quite high --- often only 15 (sometimes fewer) lines were returned, and after a while, Microsoft imposed a limit on uploading more files, even though I was deleting the conversations immediately and quitting CoPilot and re-starting (long sessions seemed to have a progressively worse failure rate).
Since then, access to CoPilot has been removed and we're waiting on a test program to finish. Kind of curious how many tokens this was using up (and if there's a model which could be run locally which would work well and quickly enough).
Replies
Enterprise sales somehow convinced a lot of orgs that all AI is the same or operates on the same level. Even frontier labs have differences, and you have to use each for your own advantage.
LLMs are not the type of model to use for this task. There are more specialized models that have much higher accuracy and full results. You can probably find the latest by looking around the OCR ecosystem.
I recently had a somewhat similar task: My 80-year old friend who is on Medicare got a $30,000 hospital bill which the insurance company declined to cover. So he wrote by hand a letter to the insurance company to request a reversal of their decision. I scanned his hand-written letter and asked AI to turn it into text (to create a professional looking letter). Could not get that to work on Gemini nor Perplexity, but it worked on Co-Pilot.
It did save me a lot time to use AI, especially should next time since I now know which AI app can handle it. So, there was a bit of overhead before I found the AI-app and prompt that would do it.
I have another friend who works for SAP as a project manager. She said she hates the fact that management wants them to use AI for everything, giving them a bonus if they do. But one reason I think their employer insists on that is that when you try to use it for real tasks you discover what works and what doesn't. Gradually we should be getting better at it.
Currently there is still a penalty caused by
1. AI not being able to do a specific task
2. Users not knowing which AI-app can do the job and
3. How to prompt the AI to get it done.
Writing the prompt is what users must do, and writing a prompt that works is a skill they must learn for every task you want AI do for you. Therefore progres is is happening but incrementally.