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

Software is mostly all you need

softwarefordays.com

Post reply on HN
Print view View original

21–30 of 52 posts

Re: Software is mostly all you need

#21

Post by kranner »

verdverm wrote: post #15
rybosworld wrote: post #10

Earlier quoted context omitted.

Neural nets have been better at classifying handwriting (MNIST) than the best humans for a long time. This is what the author means by judgement. They are super-human in their ability to classify.

Classifiers and LLMs get very different training and objectives, it's a mistake to draw inference from MNIST for coding agents or LLMs more generally. Even within coding, their capability varies widely between context and even runs with the same context. They are not better at judgement in coding for all cases, def not

A lot of the context is not even explicit, unlike the case for toy problems like MNIST.

Re: Software is mostly all you need

#22

Post by PostOnce »

rybosworld wrote: post #10
isodev wrote: post #5

> Neural networks excel at judgment I don’t think they do. I think they excel at outputting echoes of their training data that best fit (rhyme with, contextually) the prompt they were given. If you try using Claude with an obscure language or use case, you will notice that effect even more - it will keep pulling towards things it knows that aren’t at all what’s asked or “the best judgement” for what’s needed.

Neural nets have been better at classifying handwriting (MNIST) than the best humans for a long time. This is what the author means by judgement. They are super-human in their ability to classify.

Tell that to all the OCR fuckups I see in all the ebooks I read.

Re: Software is mostly all you need

#23

Post by raincole »

PostOnce wrote: post #22
rybosworld wrote: post #10

Earlier quoted context omitted.

Neural nets have been better at classifying handwriting (MNIST) than the best humans for a long time. This is what the author means by judgement. They are super-human in their ability to classify.

Tell that to all the OCR fuckups I see in all the ebooks I read.

Your ebooks are made with handwriting recognition...? What do you read, the digital version of Dead Sea Scrolls?

Re: Software is mostly all you need

#24

Post by PostOnce »

raincole wrote: post #23
PostOnce wrote: post #22

Earlier quoted context omitted.

Tell that to all the OCR fuckups I see in all the ebooks I read.

Your ebooks are made with handwriting recognition...? What do you read, the digital version of Dead Sea Scrolls?

Some of them are, most of them are standard typesetting, which you would think would be all the easier to OCR, due to the uniformity.

But because you're curious, there are some fairly famous handwritten books that maintain their handwriting in publication, my favorite being: https://boingboing.net/2020/08/31/getting-started-in-electro...

Old manuscripts are another one, there are a LOT of those. Is that handwriting? Maybe you'd argue it's "hand-printing" because its so meticulous.

Re: Software is mostly all you need

#25

Post by daxfohl »

jbmilgrom wrote: post #16
daxfohl wrote: post #14

Earlier quoted context omitted.

I agree this is what the article says, but it's a pretty bad premise. That would only be the case if the primary user interaction with coding agents was "feed in requirements, get a finished product". But we all know it's a more iterative process than that.

Author here We are building this at docflowlabs ie a self-healing system that can respond to customer feedback automatically. And youre right that not all customers know what they want or even how to express it when they do, which is why the agent loop we have facing them is way more discovery-focused than the internal one. And we currently still have humans in the loop for everything (for now!) - e.g, the agent does…

Cool, I tried something similar over a couple weeks but the problem I ran into was that beyond a fairly low level of complexity, the English spec became more confusing than the code itself. Even for a simple multi-step KYC workflow, it got very convoluted and hard to make it precise, whereas in code it's a couple loops and if/else blocks with no possibility of misinterpretation. Have you encountered that at all, or have any techniques you've found useful in these situations?

That's why I feel like iterative workflows have won out so far. Each step gets you x% closer, so you close in on your goal exponentially, whereas the one-shot approach closes in much slower, and each iteration starts from scratch. The advantage is that then you have a spec for the whole system, though you can also just generate that from the code if you write the code first.

Re: Software is mostly all you need

#26

Post by jauntywundrkind »

isodev wrote: post #5

> Neural networks excel at judgment I don’t think they do. I think they excel at outputting echoes of their training data that best fit (rhyme with, contextually) the prompt they were given. If you try using Claude with an obscure language or use case, you will notice that effect even more - it will keep pulling towards things it knows that aren’t at all what’s asked or “the best judgement” for what’s needed.

Here here. Code has uniquely an incredible volume of data. And incredibly good ways to assess & test it's weights, to immediately find out of its headed the right way on the gradient.

Re: Software is mostly all you need

#28

Post by geraneum »

jauntywundrkind wrote: post #26
isodev wrote: post #5

> Neural networks excel at judgment I don’t think they do. I think they excel at outputting echoes of their training data that best fit (rhyme with, contextually) the prompt they were given. If you try using Claude with an obscure language or use case, you will notice that effect even more - it will keep pulling towards things it knows that aren’t at all what’s asked or “the best judgement” for what’s needed.

Here here. Code has uniquely an incredible volume of data. And incredibly good ways to assess & test it's weights, to immediately find out of its headed the right way on the gradient.

> And incredibly good ways to assess & test it's weights

What weights are you referring to? How does [Claude?] code do that

Re: Software is mostly all you need

#29

Post by harry8 »

boogrpants wrote: post #17
isodev wrote: post #5

> Neural networks excel at judgment I don’t think they do. I think they excel at outputting echoes of their training data that best fit (rhyme with, contextually) the prompt they were given. If you try using Claude with an obscure language or use case, you will notice that effect even more - it will keep pulling towards things it knows that aren’t at all what’s asked or “the best judgement” for what’s needed.

> I think they excel at outputting echoes of their training data that best fit (rhyme with, contextually) the prompt they were given. Just like people who get degrees in economics or engineering and engage in such role-play for decades. They're often pretty bad at anything they are not trained on. Coincidentally, if you put a single American English speaker on a team of native German language speakers you will notice…

The fact there are a lot of people around who don't think (including me at times!) does mean LLMs doing that are thinking.

Much like LLMs writing text like mindless middle managers, it doesn't mean they're intelligent, more that mindless middle managers aren't.

Re: Software is mostly all you need

#30

Post by fooker »

jbmilgrom wrote: post #7
2001zhaozhao wrote: post #4

> Code is the policy, deployment is the episode, and the bug report is the reward signal This is a great quote. I think it makes a ton of sense to view a sufficiently-cheap-and-automated agentic SWE system as a machine learning system rather than traditional coding. * Perhaps the key to transparent/interpretable ML is to just replace the ML model with AI-coded traditional software and decision trees. This way it's st…

> Perhaps the key to transparent/interpretable ML is to just replace the ML model with AI-coded traditional software and decision trees. This way it's still fully autonomously trained but you can easily look at the code to see what is going on. For certain problems I think thats completely right. We still are not going to want that of course for classic ML domains like vision and now coding, etc. But for those domain…

> We still are not going to want that of course for classic ML domains like vision

It could make sense to decompose one large opaque model into code with decision trees calling out to smaller models having very specific purposes. This is more or less science fiction right now, 'mixture of experts' notwithstanding.

You could potentially get a Turing award by making this work for real ;)

Post reply on HN