My main use of ai.
I find the most common thing I use ai for is to give me the text from a screenshot due to websites putting important information in a manner that does not allow copy and paste of it. So I should completely not need to do this but it comes up a lot.
If you are using Windows: https://learn.microsoft.com/en-us/windows/powertoys/text-extractor
nice but I finally made the jump away but there is likely something. Not really less infuriating though since the hoops should just not need to be jumped through. I guess im mainly annoyed that the web increasingly is less text and more media.
Yeah, AI text extraction can be very powerful. It's arguably the thing that neural network systems are best at, and one of the first things. We got them to do.
There's alot of promise in using this kind of tech for quickly digitizing massive swathes of books. The more complicated part, however, is converting any images in-layed in the text.
Can't you just copy it as normal when holding alt or control (can't remember)? There is an option in firefox about:config to ignore what sites want and re-enable copy.
nope. not when they are images.
True
Man, before "AI" became a marketing gimmick,
Samsung Note devices allowed you to use the S-pen to circle text and extract it.
I remember using this feature at least a decade ago!
I think smart devices often have this but its pay for ease with computers. I have several times devled into ocr and setup was a removed. Since the chatbots do it easily it becomes my most used features. its a bit like with home assitants that I mostly ask it to set a timer.
If you are using a chromium based browser, project naphta allows you to select text in images directly in the browser.
this is working very well and saving me some mouse clicks. thanks.
There are nots of use cases for LLMs. The thing to remember is that the output can be seriously flawed and is not your property.