Substack: When you have nothing to say and all day to say it.
It was an embarrassing read. I should ask an llm to read it since he probably wrote it the same way.
A new Google model is nearly perfect on automated handwriting recognition
181–190 of 328 posts
Re: A new Google model is nearly perfect on automated handwriting recognition
#182What an unnecessarily wordy article. It could have been a fifth of the length. The actual point is buried under pages and pages of fluff and hyperbole.
I missed the point, please point me to it
Re: A new Google model is nearly perfect on automated handwriting recognition
#183Earlier quoted context omitted.
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
Sadly I'm just an amateur armchair historian (at best) so I doubt I'd be of much help. I'm mostly only doing the translation for my own edification
You may be surprised (or not?) at how many important scientific and historical works are done by armchair practitioners.
Re: A new Google model is nearly perfect on automated handwriting recognition
#184> So that is essentially the ceiling in terms of accuracy. I think this is mistaken. I remember... ten years ago? When speech-to-text models came out that dealt with background noise that made the audio sound very much like straight pink noise to my ear, but the model was able to transcribe the speech hidden within at a reasonable accuracy rate. So with handwritten text, the only prediction that makes sense to me is…
That depends on how good we get at interpretability. If the models can not only do the job but also are structured to permit an explanation of how they did it, we get the confirmation. Or not, if it turns out that the explanation is faulty.
Re: A new Google model is nearly perfect on automated handwriting recognition
#185I 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 because my entire identity is basically built around being a software engineer and thinking my IQ and intelligence is higher than other people. If this AI stuff is real then it basically destroys my entire identity so I choose the most convenient conclusion. Basically we all know that AI is just a stochastic parrot autocomplete. That's all it is. Anyone who doesn't agree with me is of lesser intelligenc…
[dead]
Re: A new Google model is nearly perfect on automated handwriting recognition
#186Earlier 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…
That is not strictly true, because being able to transfer the style of Van Gogh onto an arbitrary photographic scene is novel in a sense, but it is interpolative. Mashups are not purely derivative: the choice of what to mash up carries novelty: two (or more) representations are mashed together which hitherto have not been. We cannot deny that something is new.
Innovation itself is frequently defined as the novel combination of pre-existing components. It's mashups all the way down.
Re: A new Google model is nearly perfect on automated handwriting recognition
#187I 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 because my entire identity is basically built around being a software engineer and thinking my IQ and intelligence is higher than other people. If this AI stuff is real then it basically destroys my entire identity so I choose the most convenient conclusion. Basically we all know that AI is just a stochastic parrot autocomplete. That's all it is. Anyone who doesn't agree with me is of lesser intelligenc…
> [...] my entire identity is basically built around [...] thinking my IQ and intelligence is higher than other people.
Well, there's your first problem.
Re: A new Google model is nearly perfect on automated handwriting recognition
#188I read the whole article, but have never tried the model. Looking at the input document, I believe the model saw enough of a space between the 14 and 5 to simply treat it that way. I saw the space too. Impressive, but it's a leap to say it saw 145 then used higher order reasoning to correct 145 to 14 and 5.
I also read the whole article, and this behaviour that the author is most excited about only happened once. For a process that inherently has some randomness about it, I feel it's too early to bit this excited.
Yep. A lot of things looked magical in the GPT-4 days. Eventually you realised it did it by chance and more often than not gets it wrong
Re: A new Google model is nearly perfect on automated handwriting recognition
#189Earlier quoted context omitted.
> Predicting the next word requires understanding If we were talking about humans trying to predict next word, that would be true. There is no reason to suppose than an LLM is doing anything other than deep pattern prediction pursuant to, and no better than needed for, next word prediction.
How'd you do at the International Math Olympiad this year?
How would you do multiplying 10000 pairs of 100 digit numbers in a limited amount of time? We don't anthropomorphize calculators though...
Re: A new Google model is nearly perfect on automated handwriting recognition
#190Earlier quoted context omitted.
I hear the LLM was able to parrot fragments of the stuff it was trained to memorize, and did very well
Yeah, that must be it.
Well being able to extrapolate solutions to "novel" mathematical exercises based on a very large sample of similar tasks in your dataset seems like a reasonable explanation.
Question is how well it would do if it was trained without those samples?