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

281–290 of 328 posts

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

#281

Post by dotancohen »

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…

My language does not use Latin letters, but they are separate letters. Is there a way to train some handwriting recognition on my own handwriting in my own language, such that it will be effective and useful? I mostly need to recognize text in PDF documents, generated by writing on an e-ink tablet with an EMR stylus.

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

#282

Post by BoorishBears »

Libidinalecon wrote: post #214
BoorishBears wrote: post #154

Earlier quoted context omitted.

This is knock against you at all , but in a naive attempt to spare someone else some time: remember that based on this definition it is impossible for an LLM to do novel things and more importantly , you're not going to change how this person defines a concept as integral to one's being as novelty. I personally think this is a bit tautological of a definition, but if you hold it, then yes LLMs are not capable of anyt…

I think you should reverse the question, why would we expect LLMs to even have the ability to do novel things? It is like expecting a DJ remixing tracks to output original music. Confusing that the DJ is not actually playing the instruments on the recorded music so they can't do something new beyond the interpolation. I love DJ sets but it wouldn't be fair to the DJ to expect them to know how to play the sitar becaus…

It just depends on how you define novel.

Would you consider the instrumental at 33 seconds a new song? https://youtu.be/eJA0wY1e-zU?si=yRrDlUN2tqKpWDCv

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

#283

Post by HarHarVeryFunny »

famouswaffles wrote: post #277
HarHarVeryFunny wrote: post #272

Earlier quoted context omitted.

Regarding cats on mats ... If you ask a human to complete the phrase "the cat sat on the", they will probably answer "mat". This is memorization, not understanding. The LLM can do this too. If you just input "the cat sat on the" to an LLM, it will also likely just answer "mat" since this is what LLMs do - they are next-word input continuers. If you said "the sat sat on the" to a human, they would probably respond "hu…

>If you said "the sat sat on the" to a human, they would probably respond "huh?" or "who the hell knows!", since the human understands that cats are fickle creatures and that partial sentences are not the conversational norm. I'm not sure what you're getting at here ? You think LLMs don't similarly answer 'What are you trying to say?'. Sometimes I wonder if the people who propose these gotcha questions ever bother to…

> Sometimes I wonder if the people who propose these gotcha questions ever bother to actually test them on said LLMs

Since you asked, yes, Claude responds "mat", then asks if I want it to "continue the story".

Of course if you know anything about LLMs you should realize that they are just input continuers, and any conversational skills comes from post training. To an LLM a question is just an input whose human-preferred (as well as statistically most likely) continuation is a corresponding answer.

I'm not sure why you regard this as a "gotcha" question. If you're expressing opinions on LLMs, then table stakes should be to have a basic understanding of LLMs - what they are internally, how they work, and how they are trained, etc. If you find a description of LLMs as input-continuers in the least bit contentious then I'm sorry to say you completely fail to understand them - this is literally what they are trained to do. The only thing they are trained to do.

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

#284

Post by ChrisMarshallNY »

SJC_Hacker wrote: post #237
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.

Do you have six fingers, per chance ?

I don’t know if the six-fingered man was a Spaniard, but Inigo Montoya was…

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

#285

Post by ninetyninenine »

tptacek wrote: post #269
ninetyninenine wrote: post #222

Earlier quoted context omitted.

I am unique. Totally. It is not like HN is flooded with cognition or psychology or IQ articles every other hour. Not at all. And whenever one shows up, you do not immediately get a parade of people diagnosing themselves with whatever the headline says. Never happens. You post something about slow thinking and suddenly half the thread whispers “that is literally me.” You post something about fast thinking and the othe…

HN is not in fact flooded with cognition, psychology, and IQ articles every other hour.

There was more prior to AI but yes I exaggerated it. I mean it’s obvious right? The title of this page is hacker so it must be tech related articles every hour.

But articles on IQ and cognition and psychology are extremely common in HN. Enough to be noticeably out of place.

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

#286

Post by tptacek »

ninetyninenine wrote: post #285
tptacek wrote: post #269

Earlier quoted context omitted.

HN is not in fact flooded with cognition, psychology, and IQ articles every other hour.

There was more prior to AI but yes I exaggerated it. I mean it’s obvious right? The title of this page is hacker so it must be tech related articles every hour. But articles on IQ and cognition and psychology are extremely common in HN. Enough to be noticeably out of place.

They are actually not really all that common at all. We get 1, maybe 2 in a busy month.

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

#287

Post by famouswaffles »

HarHarVeryFunny wrote: post #283
famouswaffles wrote: post #277

Earlier quoted context omitted.

>If you said "the sat sat on the" to a human, they would probably respond "huh?" or "who the hell knows!", since the human understands that cats are fickle creatures and that partial sentences are not the conversational norm. I'm not sure what you're getting at here ? You think LLMs don't similarly answer 'What are you trying to say?'. Sometimes I wonder if the people who propose these gotcha questions ever bother to…

> Sometimes I wonder if the people who propose these gotcha questions ever bother to actually test them on said LLMs Since you asked, yes, Claude responds "mat", then asks if I want it to "continue the story". Of course if you know anything about LLMs you should realize that they are just input continuers, and any conversational skills comes from post training. To an LLM a question is just an input whose human-prefer…

Claude and GPT both ask for clarification

https://claude.ai/share/3e14f169-c35a-4eda-b933-e352661c92c2

https://chatgpt.com/share/6919021c-9ef0-800e-b127-a6c1aa8d9f...

>Of course if you know anything about LLMs you should realize that they are just input continuers, and any conversational skills comes from post training.

No, they don't. Post-training makes things easier, more accessible and consistent but conversation skills are in pre-trained LLMs just fine. Append a small transcript to the start of the prompt and you would have the same effect.

>I'm not sure why you regard this as a "gotcha" question. If you're expressing opinions on LLMs, then table stakes should be to have a basic understanding of LLMs - what they are internally, how they work, and how they are trained, etc.

You proposed a distinction and explained a situation which would make that distinction falsifiable. And I simply told you LLMs don't respond the way you claim they would. Even when models respond mat (Now I think your original point had a typo?), it is clearly not due to a lack of understanding of what normal sentences are like.

>If you find a description of LLMs as input-continuers in the least bit contentious then I'm sorry to say you completely fail to understand them - this is literally what they are trained to do. The only thing they are trained to do.

They are predictors. If the training data is solely text then the output will be more text, but that need not be the case. Words can go in while Images or actions or audio may come out. In that sense, humans are also 'input continuers'.

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

#288

Post by rtpg »

Someone wrote: post #224
yomismoaqui wrote: post #207

Earlier quoted context omitted.

I implemented a receipt scanner to Google Sheet using Gemini Flash. The fact that it is ”intelligent" it's fine for some things. For example I created structured output schema that had a field "currency" with the 3 letter format (USD, EUR...). So I scanned a receipt from some shop in Jakarta and it filled that field with IDR (Indonesian Rupiah). It inferred that data because of the city name on the receipt. Would it…

> Would it be better for my use case that it would have returned no data for the currency field? Don't think so. If there’s a decent chance it infers the wrong currency, potentially one where the value of each unit is a few units of scale larger or smaller than that of IDR, it might be better to not infer it.

I think most tools in this space do the "infer a bunch of data and show it to the user for confirmation", which lowers the pain of a miss here.

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

#289

Post by atleastoptimal »

Leaks about Gemini 3's capabilities seem to imply it is a major leap in visual acuity. I have a feeling Gemini 3 is a merger of the capabilities of Genie, Veo3 and Google's other robotics/vision developments.

However it is inevitable that people on here will try to find errors, refusing to believe the massive categorical difference in LLM's vs previous tech. The HN cycle will repeat until AGI

>HN says "AI will never/not in a long time be able to do (some arbitrary difficult task)

>Model is released that can do that task at near or above human level

>HN finds some new even narrower task (which few humans can do) that LLMs can't do yet and say that it's only "intelligent" if it can do that

>Repeat

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

#290

Post by khannn »

I have afib and have been cardioverted once ~4 years ago. Own a kardia ekg and a regular blood pressure cuff. Coffee increases my pulse, but the waveform remains normal. Alcohol is the real killer, a very small amount like a beer will immediately show in ekg waveform abnormalities. Enough to get a buzz and it's a warzone let alone actually drunk. It's also surprising how long it lasts even into the hungover and no longer drunk zone. I'm not a doctor, but I would personally recommend to cut out all drinking with an occurrence of afib and yes, this is conjecture, but any sort of heart condition. I'm also not being sponsored by kardia and am creeped out by it storing my data presumably on the cloud.
Post reply on HN