It reads to me more just as a statement of contrast, as in ‘we’re in a world of incredibly high-tech new technology, we shouldn’t still be using something from the Victorian era!’
Of cause that is a BS reason. But they should have stopped using fax machines 20 years ago. How can any reason they give why they have to stop now be any other than BS.
I was about to say, you could do serviceable OCR on a 486, which illustrates just how little processing power is needed for conventional approaches compared to this hallucinating AI nonsense.
OCR existed long before the 486. AFAIK it was already used in the 70’s or 80’s to scan mail and presort them based on the postcode. I remember that postcards had light orange boxes (presumably because this color was invisible to B/W scanners?) with dots inside where you where supposed to write the postcode numbers in.
Doing OCR in a very specific format, in a small specific area, using a set of only 9 characters, and having a list of all possible results, is not really the same problem at all.
I meant OCR of arbitrary printed or faxed text, which really only became feasible for home users in the 1990s. There were professional, but often very limited, solutions earlier than that, of course.
I don’t see how that makes sense as a statement, an ai with access to a 56k modem can send a fax. It feels like they’re just using ai as a buzzword.
It reads to me more just as a statement of contrast, as in ‘we’re in a world of incredibly high-tech new technology, we shouldn’t still be using something from the Victorian era!’
Of cause that is a BS reason. But they should have stopped using fax machines 20 years ago. How can any reason they give why they have to stop now be any other than BS.
The issue is not sending, it is receiving. With a fax you need to do some OCR to extract the text, which you then can feed into e.g an AI.
ChatGPT can recognize text on images already.
At horrendous expense, yes. Using it for OCR makes little sense. And compared to just sending the text directly, even OCR is expensive.
I was about to say, you could do serviceable OCR on a 486, which illustrates just how little processing power is needed for conventional approaches compared to this hallucinating AI nonsense.
OCR existed long before the 486. AFAIK it was already used in the 70’s or 80’s to scan mail and presort them based on the postcode. I remember that postcards had light orange boxes (presumably because this color was invisible to B/W scanners?) with dots inside where you where supposed to write the postcode numbers in.
Doing OCR in a very specific format, in a small specific area, using a set of only 9 characters, and having a list of all possible results, is not really the same problem at all.
🤔
I meant OCR of arbitrary printed or faxed text, which really only became feasible for home users in the 1990s. There were professional, but often very limited, solutions earlier than that, of course.
My phone can recognize text on images. How hard could it be to send that data to an AI?
How many billion times do you generally do that, and how is battery life after?
I wouldn’t do it on my phone. 🙄
What I’m saying is that it would probably be fairly easy to incorporate an already existing technology in to an AI.
Yes, and what I’m saying is that it would be expensive compared to not having to do it.