Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

A new Google model is nearly perfect on automated handwriting recognition

generativehistory.substack.com

Post reply on HN
Print view View original

71–80 of 328 posts

Re: A new Google model is nearly perfect on automated handwriting recognition

#71

Post by bitwize »

appreciatorBus wrote: post #25
netsharc wrote: post #7

Author says "It is the most amazing thing I have seen an LLM do, and it was unprompted, entirely accidental." and then jumps back to the "beginning of the story". Including talking about a trip to Canada. Skip to the section headed "The Ultimate Test" for the resolution of the clickbait of "the most amazing thing...". (According to him, it correctly interpreted a line in an 18th century merchant ledger using maths an…

The new model may or may not be great at handwriting but I found the author's constant repetition about how amazing it was irritating enough to stop reading and to wonder if the article itself was slop-written. "users have reported some truly wild things" "the results were shocking" "the most amazing thing I have seen an LLM do" "exciting and frightening all at once" "the most astounding result I have ever seen" "mad…

You're never gonna believe #6!

Re: A new Google model is nearly perfect on automated handwriting recognition

#72

Post by throwout4110 »

throwup238 wrote: post #61
torben-friis wrote: post #52

Earlier quoted context omitted.

>I’m interested in the Conquistadors and their extensive accounts of their expeditions, but holy cow reading 16th century handwritten Spanish and translating it at the same time is a nightmare, requiring a ton of expertise and inside field knowledge Completely off topic, but out of curiosity, where are you reading these documents? As a Spaniard I’m kinda interested.

I use the Portal de Archivos Españoles [1] for Spanish colonial documents. Each country has their own archive but the Spanish one has the most content (35 million digitized pages) The hard part is knowing where to look since most of the images haven’t gone through HRT/OCR or indexing so you have to understand Spanish colonial administration and go through the collections to find stuff. [1] https://pares.cultura.gob.e…

Want to collab on a database and some clustering and analysis? I’m a data scientist at FAIR with an interest in antiquarian docs and books

Re: A new Google model is nearly perfect on automated handwriting recognition

#73

Post by KaoruAoiShiho »

muixoozie wrote: post #65
conception wrote: post #14

I will note that 2.5 pro preview… march? Was maybe the best model I’ve used yet. The actual release model was… less. I suspect Google found the preview too expensive and optimized it down but it was interesting to see there was some hidden horsepower there. Google has always been poised to be the AI leader/winner - excited to see if this is fluff or the real deal or another preview that gets nerfed.

Dunno if you're right, but I'd like to point out that I've been reading comments like these about every model since GPT 3. It's just starting to seem more likely to me to be a cognitive bias than not.

Sometimes it is just bias but the 2.5 pro had benchmarks showing the degradation (plus they changed the name every time so it was obviously a different ckpt or model).

Re: A new Google model is nearly perfect on automated handwriting recognition

#74

Post by IAmGraydon »

orionsbelt wrote: post #47
IAmGraydon wrote: post #41

Earlier quoted context omitted.

This is utterly wrong. Predicting the next word requires a large sample of data made into a statistical model. It has nothing to do with "understanding", which implies it knows why rather than what.

Ilya Sustkever was on a podcast, saying to imagine a mystery novel where at the end it says “and the killer is: (name)”. Saying it’s just a statistical model generating the next most likely word, how can it do that in this case if it doesn’t have some understanding of all the clues, etc. A specific name is not statistically likely to appear

It can't do that without the answer to who did it being in the training data. I think the reason people keep falling for this illusion is that they can't really imagine how vast the training dataset is. In all cases where it appears to answer a question like the one you posed, it's regurgitating the answer from its training data in a way that creates an illusion of using logic to answer it.

Re: A new Google model is nearly perfect on automated handwriting recognition

#75

Post by colordrops »

muixoozie wrote: post #65
conception wrote: post #14

I will note that 2.5 pro preview… march? Was maybe the best model I’ve used yet. The actual release model was… less. I suspect Google found the preview too expensive and optimized it down but it was interesting to see there was some hidden horsepower there. Google has always been poised to be the AI leader/winner - excited to see if this is fluff or the real deal or another preview that gets nerfed.

Dunno if you're right, but I'd like to point out that I've been reading comments like these about every model since GPT 3. It's just starting to seem more likely to me to be a cognitive bias than not.

Why would you assume cognitive bias? Any evidence? These things are indeed very expensive to run, and are often run at a loss. Wouldn't quantization or other tuning be just as reasonable of an answer as cognitive bias? It's not like we are talking about reptilian aliens running the whitehouse.

Re: A new Google model is nearly perfect on automated handwriting recognition

#76

Post by snickerbockers »

jstummbillig wrote: post #40
snickerbockers wrote: post #4

Earlier quoted context omitted.

I'm skeptical that they're actually capable of making something novel. There are thousands of hobby operating systems and video game emulators on github for it to train off of so it's not particularly surprising that it can copy somebody else's homework.

I remain confused but still somewhat interested as to a definition of "novel", given how often this idea is wielded in the AI context. How is everyone so good at identifying "novel"? For example, I can't wrap my head around how a) a human could come up with a piece of writing that inarguably reads "novel" writing, while b) an AI could be guaranteed to not be able to do the same, under the same standard.

Generally novel either refers to something that is new, or a certain type of literature. If the AI is generating something functionally equivalent to a program in its training set (in this case, dozens or even hundreds of such programs) then it by definition cannot be novel.

Re: A new Google model is nearly perfect on automated handwriting recognition

#77

Post by squigz »

orionsbelt wrote: post #47
IAmGraydon wrote: post #41

Earlier quoted context omitted.

This is utterly wrong. Predicting the next word requires a large sample of data made into a statistical model. It has nothing to do with "understanding", which implies it knows why rather than what.

Ilya Sustkever was on a podcast, saying to imagine a mystery novel where at the end it says “and the killer is: (name)”. Saying it’s just a statistical model generating the next most likely word, how can it do that in this case if it doesn’t have some understanding of all the clues, etc. A specific name is not statistically likely to appear

This implies understanding of preceding tokens, no? GP was saying they have understanding of future tokens.

Re: A new Google model is nearly perfect on automated handwriting recognition

#78

Post by tanseydavid »

imiric wrote: post #64
flatline wrote: post #15

Earlier quoted context omitted.

I believe they can create a novel instance of a system from a sufficient number of relevant references - i.e. implement a set of already-known features without (much) code duplication. LLMs are certainly capable of this level of generalization due to their huge non-relevant reference set. Whether they can expand beyond that into something truly novel from a feature/functionality standpoint is a whole other, and less…

Here's a thought experiment: if modern machine learning systems existed in the early 20th century, would they have been able to produce an equivalent to the theory of relativity? How about advance our understanding of the universe? Teach us about flight dynamics and take us into space? Invent the Turing machine, Von Neumann architecture, transistors? If yes, why aren't we seeing glimpses of such genius today? If we'v…

How about "Protein Folding"?

Re: A new Google model is nearly perfect on automated handwriting recognition

#79

Post by n8cpdx »

snickerbockers wrote: post #4
throwup238 wrote: post #2

I really hope they have because I’ve also been experimenting with LLMs to automate searching through old archival handwritten documents. I’m interested in the Conquistadors and their extensive accounts of their expeditions, but holy cow reading 16th century handwritten Spanish and translating it at the same time is a nightmare, requiring a ton of expertise and inside field knowledge. It doesn’t help that they were of…

I'm skeptical that they're actually capable of making something novel. There are thousands of hobby operating systems and video game emulators on github for it to train off of so it's not particularly surprising that it can copy somebody else's homework.

The windows (~2000) kernel itself is on GitHub. Even exquisitely documented if AI can read .doc files.

https://github.com/ranni0225/WRK

Re: A new Google model is nearly perfect on automated handwriting recognition

#80

Post by imiric »

tanseydavid wrote: post #78
imiric wrote: post #64

Earlier quoted context omitted.

Here's a thought experiment: if modern machine learning systems existed in the early 20th century, would they have been able to produce an equivalent to the theory of relativity? How about advance our understanding of the universe? Teach us about flight dynamics and take us into space? Invent the Turing machine, Von Neumann architecture, transistors? If yes, why aren't we seeing glimpses of such genius today? If we'v…

How about "Protein Folding"?

A great use case for pattern recognition.
Post reply on HN