Unity's 2D UI stuff is very poorly designed, with lots of edge cases where auto-calculated fields can hit a divide-by-zero issue and then become unrecoverable because the value is now NaN which can't be auto-calculated back to a number.
You're counting just the egg-having receipts, but there were over 11 thousand receipts they had to go through to get to that 500-ish subset. I'm assuming OP wanted to process all of the receipts and then selected just eggs for a simple analytics job. With your rates, the human would cost almost $2000.
Capturing the egg price from known egg receipts was the problem I was focused on, but you're right that there was also a filtering problem in the original spec. You get my upvote for continuing to make the problem interesting for me!
Had the filtering been done during the initial document storage, then the cost would have been much cheaper than your $2,000 estimate. Essentially binning the receipts based on "eggs" or "no eggs" would be free. But, crucially, what happens when the question changes from price per egg to price per gallon of milk? Now the whole stack would need to be sorted again. The $2,000 manual classification would need to be re-applied.
Isn't traditional ML-based classification cheaper for this problem at industrial scale than an LLM though? The OP did of course attempt more traditional generic off-the-shelf OCR tools, but let's consider proper bespoke industrial ML.
Just as a off-the-cuff example, I would probably start with building a tool that locates the date/time from a receipt and takes an image snip of it. Running ONLY image snips through traditional OCR is more successful than trying to extract text from an entire receipt. I would then train a separate tool that extracts images of line items from a receipt that includes item name and price. Yet another tool could then be trained to classify items based on the names of the items purchased, and a final tool to get the price. Now you have price, item, and date to put into your database.
Perhaps generating the training data to train the item classifier is the only place I could see an LLM being more cost effective than a human, but classifying tiny image snips is not the same as one-shotting an entire receipt. As an aside, if there's any desire to discuss how expensive training ML is, don't forget the price to train an LLM as well.
All of this is to say I believe traditional ML is the solution. I'm still not seeing the value prop of LLMs at the industrialization scale outside of very targeted training data generation. A more flippant conclusion might be that we can replace a lot of the parts of data science that makes PhD types get bored with creating traditional ML solutions.
Also, playing hotdog-not-hotdog on a receipt, looking for the price of eggs, and then entering them, is a very different job than the open-ended case of "enter all the relevant information from this receipt. There is large classification task that also has to take place to group name-brand items into generic categories (an open set that you don't know from the start) suitable for analyzing.
So, I've actually done similar work to this: getting paid piece-rate to manual enter data from paper invoices into an accounting system. It was so long ago I can't remember how fast I got at it, but it was way slower than 2 a minute/120 an hour. I doubt I got much more than a dozen an hour done. So, my gut reaction is that your estimate on the human cost is off by an order of magnitude.
I don't mean to nitpick but really the only thing they have in common is a)rackets b)being played in pairs and c) being very popular at the moment. Padel is orders of magnitude richer and more complex than pickleball. Pickle is a lot of fun though.
Having played both there's a lot of overlap in the social side. You can find games with randoms or join social match plays with little friction. From what I've heard, it's not that easy with other racket sports like tennis.
I agree the skill gap is wider in padel, which can be a good thing or a bad thing, depending on who you want to play with.
I think it's important to point out that these are good manners for eating with Japanese people, not good manners for eating with chopsticks. There is no requirement to emulate Japanese eating manners if you're not in Japan and not anywhere near a person raised in Japanese cultur. There are other cultures that use chopsticks that do not necessarily have these manners.
This is definitely true - but some of these are fairly universal, or at least that is my understanding. I believe the 'no sticking chopsticks upright in rice' one is shared between Japan, Korea, China, etc. for example - it looks like funerary incense/joss sticks in all three due to the shared aspects of their cultures, for example.
I still don’t understand why making my own bowl of rice (being used/eaten by a very much alive person) look like a funeral bowl of rice is a weird or bad thing.
So much of this stuff just seems like a social license to shame people.
My mother told me, "Chuck Norris passed today at 86" and my mind immediately went to, "I would never expect him to pass anyone on the sidewalk at any slower speed."
I've gotten pretty good at identifying AI-genned music. There are two tells that I've noticed so far.
The most quantifiable is the presence of a high frequency component that sort of sounds like someone tried to clean up our restore a highly compressed track. It almost sounds kind it's going to start doing that warbling sound that happens when a teleconferencing call has a bad connection but it's just not bad enough to lose connection completely. I guess it's the sound of being highly noise gated.
The other is more qualitative. The song is boring. Like you said, on paper the song should be something I enjoy. But I suddenly notice that there is no... variation or never hook or anything to make it interesting. Anything to make it something other than the result of a machine. The aural equivalent of eating at Applebee's or reading The New Yorker. The songs just kind of plod onward without ever really getting to a point.
It feels kind of like a vivid dream when you're on the edge of lucidity. You can tell something is wrong, but there is something messing with you faculties. You're trying to see where things are going, how things will resolve, and it never happens. It just keeps going and going in a particular mode. If it does change, it's not to resolve, it's to start on a new thread that is an alternate universe version of the previous thread. With no attempt at establishing continuity, no resolution is ever found.
It had a very similar quality to the AI'd article from this thread. A sort of attempt at Being Literary but never really ever getting to the point of saying anything. It has the same feeling of wallowing, of over indulging in its shtick.
Yes, this is a thing. Bad writing with an interesting idea underneath it all is still interesting if it comes from a human because we have the expectation that the human will improve in how they share their ideas in the future. In other words, we see potential.
But LLMs don't have potential. You can make an LLM write a thousand articles in the next hour and it will not get one iota better at writing because of it. A person would massively improve merely from the act of writing a dozen, but 100x that effort and the LLM is no better off than when it started.
Despite every model release every 6 months being hailed as a "game changer", we can see from the fact that LLMs are just as empty and dumb as they were when GPT-2 was new half a decade ago that there really is no long term potential here. Despite more and more power, larger and hotter and more expensive data centers, it's an asymptotic return where we've already broken over the diminishing returns point.
And you know, I wouldn't care all that much--hell, might even be enthusiastically involved--if folks could just be honest with themselves that this turd sandwich of a product is not going to bring about AGI.
reply